Você está na página 1de 357

Mecanismos Fsicos e

Equaes de Taxas de
Transmisso de Calor
Transferncia de Calor e Energia Trmica

O que a transferncia/transmisso de calor?

A transferncia/transmisso de calor o trnsito de energia trmica


devido a uma diferena de temperaturas num meio ou entre meios.

O que a energia trmica?


A energia trmica est associada translao, rotao, vibrao e aos
estados electrnicos dos tomos e molculas que constituem a matria.

A energia trmica representa o efeito cumulativo das actividades microscpicas


e est relacionada com a temperatura da matria.
NO confundir ou trocar os significados fsicos de Energia Trmica,
Temperatura e Transferncia de Calor
Quantidade Significado fsico Smbolo Unidades
Energia Energia associada ao comportamento
Trmica+ microscpico da matria U ou u J ou J/kg
Temperatura Modo indirecto de determinar a
quantidade de energia trmica T C ou K
armazenada na matria

Transferncia de Transporte de energia trmica devido a


Calor gradientes de temperatura

Calor Quantidade de energia trmica transferida


num intervalo de tempo  t > 0
Q J
Taxa de
transferncia de
Energia trmica transferida por unidade
de tempo
q W
calor

Fluxo de calor Energia trmica transferida por unidade


de tempo e por unidade de rea
q' ' W / m2
+
U  Energia Trmica
u  Energia Trmica especfica
Modos de Transferncia de Calor

Conduo: Transferncia de calor num slido ou fluido esttico (gs ou lquido) devida ao
movimento aleatrio dos seus tomos, molculas e/ou electres constituintes.

Conveco: Transferncia de calor devida ao efeito combinado do movimento


aleatrio (microscpico) e do movimento macroscpico (adveco)
do fluido sobre uma superfcie.

Radiao: Energia que emitida pela matria devido a mudanas das configuraes
electrnicas dos seus tomos ou molculas e que transportada por ondas
electromagnticas (ou por fotes).

A conduo e a conveco exigem a presena de matria e de variaes de temperatura nesse


meio material.
Embora a radiao tenha origem na matria, o seu transporte no exige a presena de um
meio material. Alis, o transporte radiativo mais eficiente no vcuo.
Aplicaes
Identifica
Identificao de mecanismos

Problema 1.73(a): Identificao de mecanismos de transferncia de calor para janelas de vidro simples e duplo

qs Radiao solar incidente durante o dia: a fraco transmitida pelo vidro duplo menor que a transmitida pelo vidro simples.
q c o n v ,1 Conveco entre a superfcie interior da janela e o ar interior
q r a d ,1 Fluxo radiativo til trocado entre as paredes do quarto e a superfcie interior da janela
q c o n d ,1 Conduo atravs do vidro que tem superfcie interior em contacto com ar interior
q conv ,2 Conveco entre a superfcie exterior da janela e o ar exterior
q rad ,2 Fluxo radiativo til trocado entre a envolvente e a superfcie exterior da janela

q c o n d , 2 Conduo atravs do vidro que tem superfcie interior em contacto com ar exterior na janela de vidro duplo

qconv ,s Conveco no espao entre vidros (janela de vidro duplo)


qrad ,s Fluxo radiativo til entre as superfcies dos vidros que limitam o espao entre vidros
Taxas de Transferncia de Calor
Conduo
Forma geral (vectorial) da Lei de Fourier:

Fluxo de calor Condutibilidade trmica Gradiente de temperatura


W/m 2 W/m K C/m ou K/m
Aplicao ao caso de conduo unidimensional, estacionria atravs de uma
placa plana com condutibilidade trmica constante:

Fluxo de calor (W/m2):

dT T T
qx = k = k 2 1
dx L
T1 T2
qx = k
L

Taxa de transferncia de calor (W): qx = qx A


Taxas de Transferncia de Calor
Conveco

Relao entre conveco e o escoamento sobre uma superfcie e o desenvolvimento


das camadas limite hidrodinmica e trmica:

Lei do arrefecimento de Newton :

q = h (Ts T )

h [W/m2.C] ou [W/m2.K]: Coeficiente de transferncia de calor por conveco


Taxas de Transferncia de Calor
Adveco, difuso, conveco

Conveco forada, conveco natural

Calor sensvel e calor latente

Ebulio e condensao

Gama de valores tpicos do coeficiente de


conveco [W m-2 K-1]
Conveco natural - gases 2 - 25
Conveco natural - lquidos 50 - 1000
Conveco forada - gases 25 - 250
Conveco forada - lquidos 50 - 20000
Ebulio ou condensao 2500 - 100000
Taxas de Transferncia de Calor
Radiao
A transferncia de calor por radiao numa interface gs/slido envolve a emisso de
radiao a partir da superfcie e pode tambm envolver a absoro da radiao incidente
da envolvente (irradiao, G ), bem como da conveco (se Ts T)

Fluxo de energia que sai devido emisso:


E = Eb = Ts4

E [W/m2]: Poder emissivo da superfcie


(0 1): Emissividade da superfcie
Eb [W/m2]: Poder emissivo de um corpo negro (emissor perfeito)
= 5,6710-8 [W m-2 K-4] (constante de Stefan-Boltzmann)

Energia absorvida devida irradiao: Gabs = G

Gabs [W/m2]: Radiao incidente absorvida


(0 1): Absorsividade da superfcie
G [W/m2]: Irradiao
Taxas de Transferncia de Calor
Irradiao: Caso especial de uma superfcie exposta a uma
envolvente de grandes dimenses com temperatura uniforme, Tsur

G = Gsur = Tsur
4

Se = , o fluxo radiativo til a partir da superfcie


devido s trocas de calor por radiao com a envolvente :

q 'rad
'
(
= E b (TS ) G = Ts4 Tsur
4
)
Taxas de Transferncia de Calor

Em alternativa,

''
q rad = hr (TS Tsur )
[
hr W / m 2 .K ] Coeficiente de transferncia de calor por radiao

hr = (TS + Tsur )(TS2 + Tsur


2
)

Para conveco e radiao combinadas:

q = qconv = h (Ts T ) + hr (Ts Tsur )


+ qrad (1.10)
Aplicaes
Arrefecimento de componente electrnica

Problema 1.31: Dissipao de potncia em chips que operam com uma temperatura superficial de 85C
num quarto cujas paredes e ar esto a 25C para (a) conveco natural e (b) conveco forada.
Hipteses: (1) Estacionrio,
(2) Trocas de radiao entre superfcie pequena e grande
envolvente,
(3) Transferncia de calor desprezvel das faces laterais
e da superfcie de trs do chip

Pelec = q co nv + q rad = hA (Ts T ) + A (Ts4 Tsur4 )


A = L2 = ( 0.015m ) =2.2510-4 m 2
2

(a) Se for conveco natural,

qconv = CA (Ts T ) =4.2W/m 2 K 5/4 ( 2.2510-4 m 2 ) ( 60K )


5/ 4 5/4
=0.158W
qrad = 0.60 ( 2.2510-4 m 2 ) 5.6710-8 W/m 2 K 4 ( 3584 -2984 ) K 4 =0.065W
Pelec = 0.158W+0.065W=0.223W

(b) Se for conveco forada,


qconv = hA (Ts T ) =250W/m 2 K ( 2.2510-4 m 2 ) ( 60K ) =3.375W
Pelec = 3.375W+0.065W=3.44W
Conservao de Energia
CONSERVAO DE ENERGIA
(Primeira Lei da Termodinmica)
Uma ferramenta importante na anlise do fenmeno de transferncia
de calor, constituindo geralmente a base para determinar a temperatura
do sistema em estudo.

Formulaes Alternativas
Base temporal:
Num instante
ou
Num intervalo de tempo

Tipo de Sistema:
Volume de controlo
Superfcie de controlo
APLICAO A UM VOLUME DE CONTROLO
Num instante de tempo:

Notar a representao do sistema atravs de uma


superfcie de controlo (linha a tracejado) nas
fronteiras.

Fenmenos superficiais

Taxa de transferncia de energia trmica e/ou mecnica atravs da superfcie de controlo,


devido transferncia de calor, escoamento de um fluido ou transferncia de trabalho

Fenmenos volumtricos

Taxa de gerao de energia trmica devido converso de outra forma de energia (e.g.
elctrica, nuclear, qumica); converso essa de energia que ocorre no interior do sistema

Taxa de variao de energia armazenada no sistema


APLICAO A UM VOLUME DE CONTROLO
Num instante de tempo:

Notar a representao do sistema atravs de uma


superfcie de controlo (linha a tracejado line) nas
fronteiras.

Conservao de energia

Cada termo tem unidades [J/s] ou [W].

Num intervalo de tempo:

Ein + E g Eout = Est (1.11b ) Cada termo tem unidades [J].


O BALANO DE ENERGIA SUPERFICIAL
H um caso especial para o qual no existe massa ou volume contidos na superfcie de controlo

Conservao de Energia (num instante): E& in E& out = 0

Aplica-se em condies estacionrias e transientes


Sem massa nem volume, no faz sentido falar em energia armazenada ou em gerao no balano de
energia, mesmo que estes fenmenos ocorram no meio de que a superfcie faz parte.

Considere a superfcie de uma parede com transferncia de calor (conduo, conveco e radiao).

qconv
qcond qrad
= 0

k
T1 T2
L
( )
h (T2 T ) 2 T24 Tsur
4
=0
EXEMPLOS DE APLICAO
Exemplo 1.3: Aplicao resposta trmica de um fio condutor com aquecimento por efeito
de Joule (gerao de calor passagem da corrente elctrica).

E& in E& out + E& g = E& st

E& in = 0 [ (
E& out = ( D L ) h (T T ) + T 4 Tsur
4
)]
E& g = Relect I 2 E& st =
d
( c V T )
dt
EXEMPLOS DE APLICAO
Exemplo 1.43: Processamento trmico de uma bolacha de slica num forno de 2 zonas.

Sabe-se que a bolacha de slica est posicionada no forno com as superfcies


inferior e superior expostas, respectivamente, zona quente e zona fria.

Determinar
(a) Taxa inicial de aquecimento da bolacha a partir de Twi = 300K,
(b) Temperatura em regime estacionrio.
A conveco relevante?

Hipteses:
ESQUEMA a) Temperatura da bolacha uniforme
b) Temperaturas uniformes das regies quente e fria
c) Trocas radiativas entre corpo pequeno e
envolvente grande
d) Perdas da bolacha para o suporte desprezveis
EXEMPLOS DE APLICAO
Exemplo 1.43: Processamento trmico de uma bolacha de slica num forno de 2 zonas (cont)

ANLISE: No balano de energia bolacha de slica deve contabilizar-se a conveco com o gs ambiente pelas
superfcies inferior (l) e superior (u), as trocas de radiao com as zonas quente e fria e a acumulao de energia.

E& in E& out = E& st

Em termos de fluxo (por unidade de rea)


d Tw
, h + qrad
qrad , c qcv , l = cd
, u qcv
dt

( 4
Tsur 4
) 4
( 4
)
, h Tw + Tsur , c Tw hu (Tw T ) hl (Tw T ) = cd dt
w dT

(a) Como condio inicial temos Tw =Twi = 300K

( ) ( )
0.65 5.67 10 8 W / m 2 K 4 15004 3004 K 4 + 0.65 5.67 108 W / m 2 K 4 330 4 3004 K 4

8 W / m 2 K ( 300 700 ) K 4 W / m 2 K ( 300 700 ) K = 2700kg/m875J/kgK


3
0.00078 m ( d Tw / dt )i

( d Tw / dt )i = 104 K / s
EXEMPLOS DE APLICAO
Exemplo 1.43: Processamento trmico de uma bolacha de slica num forno de 2 zonas (cont)

Em regime estacionrio o armazenamento de energia nulo. O balano de energia efectuado com a temperatura
da bolacha em regime estacionrio, Tw,ss

(
0.65 15004 Tw,ss
4
) (
K 4 + 0.65 3304 Tw,ss
4
) ( ) ( )
K 4 8 W / m 2 K Tw,ss 700 K 4 W / m 2 K Tw,ss 700 K = 0

Tw,ss = 1251 K

Para determinar a importncia relativa da conveco, resolver o balano de energia sem conveco. Obtm-se
(dTw/dt)i = 101 K/s e Tw,ss = 1262 K. Logo, a radiao controla a taxa de aquecimento inicial e o regime
estacionrio.
Fouriers Law
and the
Heat Equation
Fouriers Law
A rate equation that allows determination of the conduction heat flux
from knowledge of the temperature distribution in a medium.

Its most general (vector) form for multidimensional conduction is:


r
q = k T
Implications:
Heat transfer is in the direction of decreasing temperature
(basis for minus sign).

Fouriers Law serves to define the thermal conductivity of the


medium q x
kx =
T x

Direction of heat transfer is perpendicular to lines of constant


temperature (isotherms).

Heat flux vector may be resolved into orthogonal components.


Cartesian Coordinates: T ( x, y , z )
T T T
q = k i k jk k
x y z
qx q y qz

Cylindrical Coordinates: T ( r, , z )
T T T
q = k i k jk k
r r z
qr q qz

Spherical Coordinates: T ( r , , )
T T T
q = k i k jk k
r r r sin
qr q q
In angular coordinates( or , ), the temperature gradient is still
based on temperature change over a length scale and hence has
units of C/m and not C/deg.
Heat rate for one-dimensional, radial conduction in a cylinder or sphere:

Cylinder
qr = Ar qr = 2 rLqr
or,
qr = Ar qr = 2 rqr

Sphere
qr = Ar qr = 4 r 2 qr
The Heat Equation
A differential equation whose solution provides the temperature distribution in a
stationary medium.
Based on applying conservation of energy to a differential control volume
through which energy transfer is exclusively by conduction.
Cartesian Coordinates:

T T T T
k + k + k + q = cp
x x y y z z t

Net transfer of thermal energy into the Thermal energy Change in thermal
control volume (inflow-outflow) generation energy storage
Cylindrical Coordinates:

1 T 1 T T T
kr + 2 k + k + q = c

p
r r r r z z t

Spherical Coordinates:

1 2 T 1 T 1 T T
2 r
kr +
2 2 k +
2 k sin + q = c
r sin r sin
p
r r t
One-Dimensional Conduction in a Planar Medium with Constant Properties
and No Generation

2T 1 T
=
x 2 t

k
thermal diffusivity of the medium
c p
Boundary and Initial Conditions
For transient conduction, heat equation is first order in time, requiring
specification of an initial temperature distribution: T ( x, t )t = 0 = T ( x, 0 )
Since heat equation is second order in space, two boundary conditions
must be specified. Some common cases:
Constant Surface Temperature:

T ( 0, t ) = Ts

Constant Heat Flux:


Applied Flux Insulated Surface

T T
k |x = 0 = qs |x = 0 = 0
x x

Convection

T
k |x = 0 = h T T ( 0, t )
x
Thermophysical Properties
Thermal Conductivity: A measure of a materials ability to transfer thermal
energy by conduction.

Thermal Diffusivity: A measure of a materials ability to respond to changes


in its thermal environment.

Property Tables:
Solids: Tables A.1 A.3
Gases: Table A.4
Liquids: Tables A.5 A.7
Methodology of a Conduction Analysis
Solve appropriate form of heat equation to obtain the temperature
distribution.

Knowing the temperature distribution, apply Fouriers Law to obtain the


heat flux at any time, location and direction of interest.

Applications:

Chapter 3: One-Dimensional, Steady-State Conduction


Chapter 4: Two-Dimensional, Steady-State Conduction
Chapter 5: Transient Conduction
Problem 2.46 Thermal response of a plane wall to convection heat transfer.

KNOWN: Plane wall, initially at a uniform temperature, is suddenly exposed to convective heating.

FIND: (a) Differential equation and initial and boundary conditions which may be used to find the
temperature distribution, T(x,t); (b) Sketch T(x,t) for the following conditions: initial (t 0), steady-
state (t ), and two intermediate times; (c) Sketch heat fluxes as a function of time at the two
surfaces; (d) Expression for total energy transferred to wall per unit volume (J/m3).

SCHEMATIC:
ASSUMPTIONS: (1) One-dimensional conduction, (2) Constant properties, (3) No internal
heat generation.
ANALYSIS: (a) For one-dimensional conduction with constant properties, the heat equation has the
form,

2T 1 T
=
x2 t

Initial, t 0 : T ( x,0 ) = Ti uniform temperature



and the Boundaries: x=0 T/ x)0 = 0 adiabatic surface

conditions are:
x=L k T/ x) L = h T ( L,t ) T surface convection

(b) The temperature distributions are shown on the sketch.

Note that the gradient at x = 0 is always zero, since this boundary is adiabatic. Note also that the
gradient at x = L decreases with time.
c) The heat flux, q x ( x, t ) as a function of time, is shown on the sketch for the surfaces x = 0 and
x = L.

d) The total energy transferred to the wall may be expressed as


Ein = qconv As dt
0

Ein = hAs
0
( T T ( L,t ) )dt

Dividing both sides by AsL, the energy transferred per unit volume is

Ein h
T T ( L,t ) dt J/m3
V L 0
=

Problem: Non-
Non-uniform Generation due
to Radiation Absorption

Problem 2.28 Surface heat fluxes, heat generation and total rate of radiation
absorption in an irradiated semi-transparent material with a
prescribed temperature distribution.

KNOWN: Temperature distribution in a semi-transparent medium subjected to radiative flux

FIND: (a) Expressions for the heat flux at the front and rear surfaces, (b) The heat generation rate
q& ( x ) , and (c) Expression for absorbed radiation per unit surface area.

SCHEMATIC :
Problem : Non-
Non-uniform
Generation (Cont.)

ASSUMPTIONS: (1) Steady-state conditions, (2) One-dimensional conduction in medium, (3)


Constant properties, (4) All laser irradiation is absorbed and can be characterized by an internal
volumetric heat generation term q& ( x ) .

ANALYSIS: (a) Knowing the temperature distribution, the surface heat fluxes are found using
Fouriers law,
dT A
q x = k = k - 2 ( a ) e-ax + B
dx ka
A A
Front Surface, x=0: q x ( 0 ) = k + 1 + B = + kB <
ka a
A A
Rear Surface, x=L: q x ( L ) = k + e-aL + B = e-aL + kB . <
ka a

(b) The heat diffusion equation for the medium is


d dT q& d dT
+ =0 or &
q=-k
dx dx k dx dx
d A -ax
q& ( x ) = k + e + B = Ae -ax
.
dx ka

( c ) Performing an energy balance on the medium,


E& in E& out + E& g = 0
Problem : Non-
Non-uniform
Generation (Cont.)

On a unit area basis

E& g = E& in
+ E& out ( A
a
)
= q x ( 0 ) + q x ( L ) = + 1 e-aL . <

& by integration over the volume of the medium,


Alternatively, evaluate E g

( )
A L A
E& g = q& ( x )dx= Ae-ax dx=- e-ax =
L L
1 e-aL .
0 0 a 0 a
One-Dimensional, Steady-State
Conduction without
Thermal Energy Generation
Methodology of a Conduction Analysis
Specify appropriate form of the heat equation.
Solve for the temperature distribution.
Apply Fouriers Law to determine the heat flux.

Simplest Case: One-Dimensional, Steady-State Conduction with No Thermal Energy Generation

Alternative conduction analysis

Common Geometries:
The Plane Wall: Described in rectangular (x) coordinate. Area
perpendicular to direction of heat transfer is constant (independent of x).
The Tube Wall: Radial conduction through tube wall.
The Spherical Shell: Radial conduction through shell wall.
The Plane Wall
Consider a plane wall between two fluids of different temperature:

Heat Equation:
d dT
k =0
dx dx

Implications:
Heat flux ( qx ) is independent of x.
Heat rate ( q x ) is independent of x.
Boundary Conditions: T ( 0 ) = Ts ,1, T ( L ) = Ts ,2

Temperature Distribution for Constant k :


T ( x ) = Ts ,1 + (Ts ,2 Ts ,1 )
x
L
Heat Flux and Heat Rate:
= (Ts ,1 Ts ,2 )
dT k
qx = k
dx L
q x = kA
dT kA
dx
=
L
( Ts ,1 Ts ,2 )

T
R
Thermal Resistances t = and Thermal Circuits:
q
L
Conduction in a plane wall: Rt , cond =
kA
1
Convection: Rt ,conv =
hA
Thermal circuit for plane wall with adjoining fluids:

1 L 1
Rtot = + +
h1 A kA h 2 A
T,1 T,2
qx =
Rtot
Thermal Resistance for Unit Surface Area:
L 1
Rt,cond = Rt,conv =
k h
Units: Rt W/K Rt m 2 K/W
Radiation Resistance:
1 1
Rt , rad = Rt, rad =
hr A hr
(
hr = (Ts + Tsur ) Ts2 + Tsur
2
)
Contact Resistance:

TA TB Rt,c
Rt,c = Rt ,c =
qx Ac

Values depend on: Materials A and B, surface finishes, interstitial conditions, and
contact pressure (Tables 3.1 and 3.2)
Composite Wall with Negligible Contact Resistance:

T,1 T,4
qx =
Rtot

1 1 LA LB LC 1 Rtot
Rtot = + + + + =
A h1 k A k B kC h4 A

Overall Heat Transfer Coefficient (U) :


A modified form of Newtons Law of Cooling to encompass multiple resistances
to heat transfer.
q x = UAToverall
1
Rtot =
UA
Series Parallel Composite Wall:

Note departure from one-dimensional conditions for k F kG .

Circuits based on assumption of isothermal surfaces normal to x direction or


adiabatic surfaces parallel to x direction provide approximations for q x .
ALTERNATIVE CONDUCTION ANALYSIS:

STEADY STATE
NO HEAT GENERATION
NO HEAT LOSS FROM THE SIDES
A(x) and k(T)

IS TEMPERATURE DISTRIBUTION ONE-DIMENSIONAL? q x = q x + dx


IS IT REASONABLE TO ASSUME ONE-DIMENSIONAL
TEMPERATURE DISTRIBUTION IN x?

dT
FROM THE FOURIERS LAW: q x = A( x ) k (T )
dx
x T
dx
qx = k (T )dT
0
A( x) T0
Tube Wall
The Tube Wall

Heat Equation:
1 d dT
kr =0
r dr dr
What does the form of the heat equation tell us about the variation of qr with
r in the wall?
Is the foregoing conclusion consistent with the energy conservation requirement?

How does qr vary with r ?


Temperature Distribution for Constant k:
Ts ,1 Ts ,2 r
T (r ) = ln + Ts ,2
ln ( r1 / r2 ) r2
Heat Flux and Heat Rate:

qr = k
dT
=
k
dr r ln ( r2 / r1 )
(Ts,1 Ts,2 )
2 k
qr = 2 rqr =
ln ( r2 / r1 )
( Ts ,1 Ts ,2 )

2 Lk
qr = 2 rLqr =
ln ( r2 / r1 )
( Ts ,1 Ts ,2 ) (3.27)

Conduction Resistance:
ln ( r2 / r1 )
Rt ,cond = Units K/W
2 Lk
ln ( r2 / r1 )
Rt,cond = Units m K/W
2 k

Why is it inappropriate to base the thermal resistance on a unit surface


area?
Composite Wall with
Negligible Contact
Resistance

T,1 T,4
qr = = UA (T,1 T ,4 )
Rtot

Note that
UA = Rtot 1
is a constant independent of radius.

But, U itself is tied to specification of an interface.


1
U i = ( Ai Rtot )
Spherical Shell

Heat Equation
1 d 2 dT
2 dr
r =0
r dr
What does the form of the heat equation tell us about the variation of
qr with r ? Is this result consistent with conservation of energy?

How does qr vary with r ?

Temperature Distribution for Constant k :

T ( r ) = Ts ,1 (Ts ,1 Ts ,2 )
( )
1 r1/ r
(
1 r1 / r 2 )
Heat flux, Heat Rate and Thermal Resistance:
dT
qr = k = 2
k
dr r (1/ r1 ) (1/ r2 )
(Ts,1 Ts,2 )
4 k
qr = 4 r 2 qr =
( 1) ( 2 )
1/ r 1/ r
( Ts ,1 Ts ,2 )

Rt ,cond =
(1/ r1 ) (1/ r2 )
4 k

Composite Shell:
T
qr = overall = UAToverall
Rtot
UA = Rtot 1 Constant
1
U i = ( Ai Rtot ) Depends on Ai
r1 r2
r T ,h
Critical radius (cylindrical geometry)
T,1 ,h1
Isolamento

(a)

T,1 T

1 ln (r2 /r1 ) ln (r / r2 ) 1
2p r1 L h1 2p k1 L 2p k L 2p r L h
T,1 T (b)
q r ,sem revest . =
1 ln (r2 r1 ) 1
+ +
2 r1 L h1 2 k1 L 2 r2 L h

T,1 T
q r ,com revest . =
1 ln(r2 r1 ) ln (r r2 ) 1
+ + +
2 r1 L h1 2 k1 L 2 k L 2 r L h

d Rtot 1 1 1 1
=
dr 2 k L r 2 L h r 2

d Rtot k d 2 Rtot 1 1 1 2
=0 rcrit =
d r2


= 2 + 3 >0
h r =k 2 k L r 2 L h r r =k
dr h h
Problem 3.23: Assessment of thermal barrier coating (TBC) for protection
of turbine blades. Determine maximum blade temperature
with and without TBC.

Schematic:

ASSUMPTIONS: (1) One-dimensional, steady-state conduction in a composite plane wall, (2) Constant
properties, (3) Negligible radiation
ANALYSIS: For a unit area, the total thermal resistance with the TBC is

R tot, w = h o1 + ( L k ) Zr + R t,c + ( L k )In + h i1

( )
R tot, w = 103 + 3.85 104 + 104 + 2 104 + 2 103 m 2 K W = 3.69 10 3 m 2 K W

With a heat flux of


T,o T,i 1300 K
q w = = = 3.52 105 W m 2
R tot, w 3.69 103 m 2 K W

the inner and outer surface temperatures of the Inconel are

( )
Ts,i(w) = T,i + ( q w h i ) = 400 K + 3.52 105 W m 2 500 W m 2 K = 1104 K

( ) ( )
Ts,o(w) = T ,i + (1 h i ) + ( L k ) In q w = 400 K + 2 10 3 + 2 10 4 m 2 K W 3.52 105 W m 2 = 1174 K
Without the TBC,

1 1 3
R tot, wo = h o + ( L k )In + h i
2
= 3.20 10 m K W

(
q wo = T ,o T ,i ) R tot, wo = 4.06105 W/m2

The inner and outer surface temperatures of the Inconel are then

Ts,i(wo) = T,i + ( q wo h i ) = 1212 K

Ts, o(wo) = T ,i + [ (1 h i ) + ( L ]
k ) In q wo
= 1293 K

Use of the TBC facilitates operation of the Inconel below Tmax = 1250 K.

COMMENTS: Since the durability of the TBC decreases with increasing


temperature, which increases with increasing thickness, limits to its thickness are
associated with reliability considerations.
Problem 3.62: Suitability of a composite spherical shell for storing
radioactive wastes in oceanic waters.

SCHEMATIC:

ASSUMPTIONS: (1) One-dimensional conduction, (2) Steady-state conditions,


(3) Constant properties at 300K, (4) Negligible contact resistance.

PROPERTIES: Table A-1, Lead: k = 35.3 W/mK, MP = 601K; St.St.: 15.1


W/mK.

ANALYSIS: From the thermal circuit, it follows that


T T 4
q= 1 = q& r13
R tot 3
The thermal resistances are:
1 1
R Pb = 1/ ( 4 35.3 W/m K ) = 0.00150 K/W
0.25m 0.30m
1 1
R St.St. = 1/ ( 4 15.1 W/m K ) = 0.000567 K/W
0.30m 0.31m

( )
R conv = 1/ 4 0.312 m 2 500 W/m 2 K = 0.00166 K/W

R tot = 0.00372 K/W.

The heat rate is then


q=5 105 W/m3 ( 4 / 3)( 0.25m ) = 32, 725 W
3

and the inner surface temperature is


T1 = T + R tot q=283K+0.00372K/W ( 32,725 W ) = 405 K < MP = 601K.

Hence, from the thermal standpoint, the proposal is adequate.

COMMENTS: In fabrication, attention should be given to maintaining a good


thermal contact. A protective outer coating should be applied to prevent long
term corrosion of the stainless steel.
One-Dimensional, Steady-State
Conduction with
Thermal Energy Generation
Implications of Energy Generation

Involves a local (volumetric) source of thermal energy due to conversion


from another form of energy in a conducting medium.

The source may be uniformly distributed, as in the conversion from


electrical to thermal energy (Ohmic heating):

E& g I2 R
q& = =
V V

or it may be non-uniformly distributed, as in the absorption of radiation


passing through a semi-transparent medium. For a plane wall,

q& exp( x )

Generation affects the temperature distribution in the medium and causes


the heat rate to vary with location, thereby precluding inclusion of
the medium in a thermal circuit.
The Plane Wall
Consider one-dimensional, steady-state conduction
in a plane wall of constant k, uniform generation,
and asymmetric surface conditions:

Heat Equation:

d dT 2
dT q
k + q = 0 + =0 (3.39)
dx dx dx 2 k

Is the heat flux q independent of x?

General Solution:

T ( x ) = q/ 2k x 2 + C1 x + C2

What is the form of the temperature distribution for



q = 0? q > 0? q < 0?


How does the temperature distribution change with increasing q ?
Symmetric Surface Conditions or One Surface Insulated:

What is the temperature gradient


at the centerline or the insulated
surface?
Why does the magnitude of the temperature
gradient increase with increasing x?

Temperature Distribution:

q L2 x 2
T ( x) = 1 + Ts
2k L2
(3.42)

How do we determine Ts ?
Overall energy balance on the wall

E out + E g = 0

hAs (Ts T ) + q As L = 0


qL
Ts = T + (3.46)
h
How do we determine the heat rate at x = L?
Radial Systems
Cylindrical (Tube) Wall Spherical Wall (Shell)

Solid Cylinder (Circular Rod) Solid Sphere

Heat Equations:
Cylindrical Spherical
1 d dT 1 d 2 dT
kr +q =0 kr +q =0
r dr dr r 2 dr dr
Solution for Uniform Generation in a Solid Sphere of Constant k
with Convection Cooling:

Temperature Distribution Surface Temperature



dT q r3 Overall energy balance:
kr2
= + C1
dr 3

2 qr
T =
q r C1
+ C2 E out + Eg = 0 Ts = T + o
6k r 3h
dT
|r = 0 = 0 C1 = 0
dr Or from a surface energy balance:

2
q ro
T ( ro ) = Ts C2 = Ts +
6k q ro
E in E out = 0 qcond ( ro ) = qconv Ts = T +


q ro 2
r 2

T (r ) = 1 2 + Ts
3h
6k ro

A summary of temperature distributions is provided in Appendix C


for plane, cylindrical and spherical walls, as well as for solid
cylinders and spheres. Note how boundary conditions are specified
and how they are used to obtain surface temperatures.
Problem 3.91 Thermal conditions in a gas-cooled nuclear reactor
with a tubular thorium fuel rod and a concentric
graphite sheath: (a) Assessment of thermal integrity
for a generation rate of q = 108 W/m 3. (b) Evaluation of
temperature distributions in the thorium
and graphite
for generation rates in the range 108 q 5x108.

Schematic:

Assumptions: (1) Steady-state conditions, (2) One-dimensional conduction, (3) Constant


properties, (4) Negligible contact resistance, (5) Negligible radiation, (6) Adiabatic surface at r1.

Properties: Table A.1, Thorium: Tmp 2000 K ; Table A.2, Graphite: Tmp 2300 K .
Analysis: (a) The outer surface temperature of the fuel, T2 , may be determined from the rate equation

T2 T
q =

Rtot
1n ( r3 / r2 ) 1
=
where Rtot + = 0.0185 m K/W
2 k g 2 r3h

The heat rate may be determined by applying an energy balance to a control surface about the fuel
element,
E out = E g

or, per unit length, E out = E g

Since the interior surface of the element is essentially adiabatic, it follows that

q = q ( r22 r12 ) = 17,907 W/m


Hence,
+ T = 17,907 W/m ( 0.0185 m K/W ) + 600 K = 931K
T2 = qRtot
With zero heat flux at the inner surface of the fuel element, Eq. C.14 yields

2

q r r1 q r12 r2
2

T1 = T2 + 2
1 1n = 931K + 25 K 18 K = 938 K <
4kt r22 2kt r1

Since T1 and T2 are well below the melting points of thorium and graphite, the prescribed
operating condition is acceptable.

(b) The solution for the temperature distribution in a cylindrical wall with generation is

q r22 r 2
Tt ( r ) = T2 + 1 2
4kt r2
2
q r r 2
1n ( r2 / r )
2 1 12 + (T2 T1 ) (C.2)
4kt r 1n ( r2 / r1 )


2

Boundary conditions at r1 and r2 are used to determine T1 and T 2 .

q r 2 r 2
k 1 2 + (T2 T1 )
2 1
(C.14)
4kt r2

r = r1 : q1 = 0 =
qr1

2 r11n ( r2 / r1 )

qr 2 r 2
k 1 2 + (T2 T1 )
2 1 (C.17)

4kt r2
r = r2 : U 2 (T2 T ) =
q r2

2 r21n ( r2 / r1 )

1 1
U 2 = ( A2 Rtot
) = ( 2 r2 Rtot
) (3.32)
The following results are obtained for temperature distributions in the graphite.
2500

2100
Temperature, T(K)

1700

1300

900

500
0.008 0.009 0.01 0.011
Radial location in fuel, r(m)

qdot = 5E8
qdot = 3E8
qdot = 1E8


Operation at q = 5x108 W/m 3 is clearly unacceptable since the melting point of
thorium would be exceeded. To prevent softening of the material, which would occur

below the melting point, the reactor should not be operated much above q = 3x10 W/m.
8 3

The small radial temperature gradients are attributable to the large value of kt .
Using the value of T2 from the foregoing solution and computing T3 from the surface condition,
2 k g (T2 T3 )
q = (3.27)
1n ( r3 / r2 )

the temperature distribution in the graphite is


T2 T3 r
Tg ( r ) = 1n + T3 (3.26)
1n ( r2 / r3 ) r3

2500

2100
Temperature, T(K)

1700

1300

900

500
0.011 0.012 0.013 0.014
Radial location in graphite, r(m)

qdot = 5E8
qdot = 3E8
qdot = 1E8

Operation at q = 5x108 W/m 3 is problematic for the graphite. Larger temperature gradients
are due to the small value of k g .
Comments: (i) What effect would a contact resistance at the thorium/graphite

interface have on
temperatures in the fuel element and on the maximum allowable value of q ? (ii) Referring
to the schematic, where might radiation effects be significant? What would be the influence of such

effect on temperatures in the fuel element and the maximum allowable value of q ?
Extended Surfaces
Nature and Rationale of Extended Surfaces
An extended surface (also know as a combined conduction-convection system
or a fin) is a solid within which heat transfer by conduction is assumed to be
one dimensional, while heat is also transferred by convection (and/or
radiation) from the surface in a direction transverse to that of conduction.

Why is heat transfer by conduction in the x-direction not, in fact, one-


dimensional?

If heat is transferred from the surface to the fluid by convection, what


surface condition is dictated by the conservation of energy requirement?
What is the actual functional dependence of the temperature distribution in
the solid?
If the temperature distribution is assumed to be one-dimensional, that is,
T=T(x) , how should the value of T be interpreted for any x location?
How does qcond , x vary with x ?
When may the assumption of one-dimensional conduction be viewed as an
excellent approximation? The thin-fin approximation.
Extended surfaces may exist in many situations but are commonly used as
fins to enhance heat transfer by increasing the surface area available for
convection (and/or radiation). They are particularly beneficial when h is small,
as for a gas and natural convection.
Some typical fin configurations:

Straight fins of (a) uniform and (b) non-uniform cross sections; (c) annular
fin, and (d) pin fin of non-uniform cross section.
TYPICAL FIN CONFIGURATIONS

(a) (b) (c)

(d) (e) (f)

(g) (h) (i)


The Fin Equation
dqconv
dAs
qx
Ac (x)
q x = q x + dx + d qconv
qx+dx

dx

z x
y
x

dT d qx dT d dT
q x = k Ac q x+ dx = q x + dx = k Ac k Ac dx
dx dx dx d x d x

d qconv = h dAs (T T )

d dT d As
k Ac + h (T T ) = 0
d x d x dx

d2T 1 d Ac d T 1 h d As
2
+ (T T ) = 0
dx Ac d x d x Ac k d x
The Fin Equation
Assuming one-dimensional, steady-state conduction in an extended surface
surface of constant conductivity ( k ) and uniform cross-sectional area ( Ac,)
with negligible generation q = 0 and radiation ( qrad
= 0 ) , the fin equation

is of the form:

d 2T hP
2
(T T ) = 0 (3.62)
dx kAc

or, with m2 ( hP / kAc ) and the reduced temperature T T ,

d 2
2
m 2
=0
dx
Solutions (Table 3.4):

Base (x = 0) condition
( 0 ) = Tb T b

Tip ( x = L) conditions
A. Convection: kd / dx |x = L = h ( L )
B. Adiabatic: d / dx |x = L = 0
C. Fixed temperature: ( L ) = L
D. Infinite fin (mL >2.65): ( L ) = 0

Fin Heat Rate:


d
q f = kAc |x = 0 = Af h ( x ) dAs
dx
Condio de Distribuio de temperaturas Taxa de transmisso de
Caso
fronteira em x = L /b calor
cosh[m(L x )] + sinh[m(L x )] sinh (m L ) + cosh (m L )
h h
d
= h (L )
mk mk
(i) k M
cosh (m L ) + sinh (m L ) cosh (m L ) + sinh (m L )
d x x=L h h
mk mk
d cosh [m(L x )]
(ii) =0 M tanh (m L )
d x x =L cosh (m L )

( L b ) sinh (m x ) + sinh [m(L x )] cosh (m L ) L / b


(iii) (L ) = L M
sinh (m L ) sinh (m L )

(iv) (L ) = 0 em x M

hP
m2 = M = h P k Ac b
k Ac
Fin Performance Parameters
Fin Efficiency:
qf qf
f =
q f ,max hA f b
How is the efficiency affected by the thermal conductivity of the fin?
Expressions for f are provided in Table 3.5 for common geometries.
Consider a triangular fin:
1/ 2
A f = 2 w L2 + ( t / 2 )
2

Ap = ( t / 2 ) L
1 I1 ( 2mL )
f =
mL I 0 ( 2mL )

Fin Effectiveness:
qf
f
hAc , bb
f with h, k and Ac / P
Fin Resistance:
b 1
Rt , f =
qf hA f f
Correction of fin length to account for heat loss from the tip

Transmisso de calor
q f ,tip = h Ac (L ) h P (Lc L ) (L )
na extremidade

Ac
Lc = L +
P

Fin of rectangular cross section with t << w:

Lc = L + t / 2

extremidade
Fin of circular cross section :
isolada

Lc = L + D / 4

Approximation error negligible if ht / k or


hD / 2k 0.0625
Fins efficiency

1.0 1.0 y (x)


(a) t
ri
0.9
x
(b)
0.8 L 0.8

t
0.7 (c)

0.6 0.6
ro
(d)
hf hf 0.5
ro
=1 (e)
0.4
ri
0.4

1.4 0.3

0.2 1.6 0.2


3 1.8
2
4 0.1

0.0 0.0
0.0 1.0 2.0 3.0 4.0 5.0 0.0 1.0 2.0 3.0 4.0 5.0
Fin Arrays
Representative arrays of
(a) rectangular and
(b) annular fins.

Total surface area:


At = NA f + Ab
Number of fins Area of exposed base (prime surface)

Total heat rate:


b
qt = N f hA f b + hAbb o hAtb =
Rt , o
Overall surface efficiency and resistance:
(1 f )
NA f
o = 1
At
b 1
Rt , o = =
qt o hAt
Equivalent Thermal Circuit :

Effect of Surface Contact Resistance:

b
qt = o ( c ) hAtb =
Rt , o ( c )
NA f f
o ( c ) = 1 1
At C1
C1 = 1 + f hA f ( Rt, c / Ac , b )
1
Rt , o ( c ) =
o ( c ) hAt
Problem 3.116: Assessment of cooling scheme for gas turbine blade.
Determination of whether blade temperatures are less
than the maximum allowable value (1050 C) for
prescribed operating conditions and evaluation of blade
cooling rate.

Schematic:

Assumptions: (1) One-dimensional, steady-state conduction in blade, (2) Constant k, (3)


Adiabatic blade tip, (4) Negligible radiation.

Analysis: Conditions in the blade are determined by Case B of Table 3.4.

(a) With the maximum temperature existing at x=L, Eq. 3.75 yields

T ( L ) T 1
=
Tb T cosh mL

( )
1/ 2
m = ( hP/kA c ) = 250W/m 2 K 0.11m/20W/m K 6 104 m 2
1/ 2
= 47.87 m-1

mL = 47.87 m-1 0.05 m = 2.39


From Table B.1, coshmL=5.51 . Hence,

T ( L ) = 1200o C + (300 1200)o C/5.51 = 1037 o C

and, subject to the assumption of an adiabatic tip, the operating conditions are acceptable.

( ) ( 900 C ) = 517W ,
1/ 2
(b) With M = ( hPkA c )1 / 2 b = 250W/m 2 K 0.11m 20W/m K 6 10 4 m 2 o

Eq. 3.76 and Table B.1 yield

q f = M tanh mL = 517W ( 0.983) = 508W

Hence, q b = q f = 508W

Comments: Radiation losses from the blade surface contribute to reducing the blade
temperatures, but what is the effect of assuming an adiabatic tip condition? Calculate
the tip temperature allowing for convection from the gas.
Problem 3.132: Determination of maximum allowable power qc for a 20mm
x 20mm electronic chip whose temperature is not to exceed
Tc = 85o C, when the chip is attached to an air-cooled heat sink
with N=11 fins of prescribed dimensions.

Schematic:

Assumptions: (1) Steady-state, (2) One-dimensional heat transfer, (3) Isothermal chip, (4)
Negligible heat transfer from top surface of chip, (5) Negligible temperature rise for air flow,
(6) Uniform convection coefficient associated with air flow through channels and over outer
surface of heat sink, (7) Negligible radiation.
Analysis: (a) From the thermal circuit,

T T Tc T
qc = c =
R tot R t,c + R t,b + R t,o
6 2
R t,c = R t,c / W = 2 10 m K / W / ( 0.02m ) = 0.005 K / W
2 2

R t, b = L b / k W ( ) = 0.003m / 180
2
W/mK ( 0.02m )2 = 0.042 K / W

1 N Af
From Eqs. (3.103), (3.102), and (3.99) R t,o = , o = 1 (1 f ) , A t = N Af + A b
o h A t At

-4 2
Af = 2WLf = 2 0.02m 0.015m = 6 10 m
2 2 -3 -4 2
Ab = W N(tW) = (0.02m) 11(0.182 10 m 0.02m) = 3.6 10 m
-3 2
At = 6.96 10 m
1/2 2 -3 1/2
With mLf = (2h/kt) Lf = (200 W/m K/180 W/mK 0.182 10 m) (0.015m) =
1.17, tanh mLf = 0.824 and Eq. (3.87) yields
tanh mLf 0.824
f = = = 0.704
mLf 1.17

o = 0.719,
Rt,o = 2.00 K/W, and

(85 20 ) C
qc = = 31.8 W
( 0.005 + 0.042 + 2.00 ) K / W
Comments: The heat sink significantly increases the allowable heat dissipation. If it
were not used and heat was simply transferred by convection from the surface of the chip with
h = 100 W/m 2 K , Rtot = 2.05 K/W from Part (a) would be replaced by
Rcnv = 1/ hW 2 = 25 K/W, yielding qc = 2.60 W.
Transient Conduction:
The Lumped Capacitance
Method
Transient Conduction
A heat transfer process for which the temperature varies with time, as well
as location within a solid.

It is initiated whenever a system experiences a change in operating conditions


and proceeds until a new steady state (thermal equilibrium) is achieved.

It can be induced by changes in:


surface convection conditions ( h, T ),
surface radiation conditions ( hr , Tsur ),
a surface temperature or heat flux, and/or
internal energy generation.

Solution Techniques
The Lumped Capacitance Method
Exact Solutions
The Finite-Difference Method (not to be studied)
The Lumped Capacitance Method
Based on the assumption of a spatially uniform temperature distribution
r
throughout the transient process.
T ( r ,t ) T ( t )
Why is the assumption never fully realized in practice?

General Lumped Capacitance


Analysis:

 Consider a general case,


which includes convection,
radiation and/or an applied
heat flux at specified
surfaces ( As , c , As , r , As , h ) ,
as well as internal energy
generation
 First Law:

d E st dT
= V C = E& in E& out + E& g
dt dt
Assuming energy outflow due to convection and radiation and with
inflow due to an applied heat flux qs,

dT
V C = q's',h As ,h hAs ,c ( T T ) hs ,r As ,r ( T Tsur ) + E& g
dt

Is this expression applicable in situations for which convection and/or


radiation provide for energy inflow?

May h and hr be assumed to be constant throughout the transient process?

How must such an equation be solved?


Special Cases (Exact Solutions, T ( 0 ) Ti )
 Negligible Radiation ( T T , b / a ) :
h As ,c q '' As , h + E& g
a= b=
V C V C
The non-homogeneous differential equation is transformed into a
homogeneous equation of the form:
d
= a
dt

Integrating from t=0 to any t and rearranging,


T T b/a
= exp ( at ) + 1 exp ( at )
Ti T Ti T

To what does the foregoing equation reduce as steady state is approached?


How else may the steady-state solution be obtained?
 Negligible Radiation and Source Terms h >> hr , Eg = 0, qs = 0 :
dT
c = hAs , c (T T )
dt
c d t

hAs , c i
= dt
o
T T hAs , c t
= = exp
t = exp
i Ti T c t

The thermal time constant is defined as


1
t ( c )
hAs , c
Thermal Lumped Thermal
Resistance, Rt Capacitance, Ct

The change in thermal energy storage due to the transient process is


t t t
Est Q = Eout dt = hAs , c dt = ( c )i 1 exp (5.8)
o o t
h >> h, E = 0, q = 0 :
 Negligible Convection and Source Terms r g s

Assuming radiation exchange with large surroundings,


dT
c = As , r (T 4 Tsur
4
)
dt
A s , r t T dT
=
c o T i T 4 T
dt 4
sur

c Tsur + T Tsur + Ti
t= 1n 1n
4 As , r Tsur
3
Tsur T Tsur Ti

T Ti
+2 tan 1 tan 1

T sur T sur

Result necessitates implicit evaluation of T(t).


The Biot Number and Validity of
The Lumped Capacitance Method
The Biot Number: The first of many dimensionless parameters to be
considered.
 Definition:
hL
Bi c
k
h convection or radiation coefficient
k thermal conductivity of the solid
Lc characteristic length of the solid ( / As or coordinate
associated with maximum spatial temperature difference)

 Physical Interpretation:

Lc / kAs Rcond Tsolid


Bi = = =
1/ hAs Rconv Tsolid / fluid

 Criterion for Applicability of Lumped Capacitance Method:


Bi << 1
Problem 5.11: Charging a thermal energy storage system consisting
of a packed bed of aluminum spheres.

KNOWN: Diameter, density, specific heat and thermal conductivity of aluminum spheres used in
packed bed thermal energy storage system. Convection coefficient and inlet gas temperature.

FIND: Time required for sphere at inlet to acquire 90% of maximum possible thermal energy and the
corresponding center temperature.

Schematic:
ASSUMPTIONS: (1) Negligible heat transfer to or from a sphere by radiation or conduction due to
contact with other spheres, (2) Constant properties.
ANALYSIS: To determine whether a lumped capacitance analysis can be used, first compute Bi =
2
h(ro/3)/k = 75 W/m K (0.025m)/150 W/mK = 0.013 <<1.
Hence, the lumped capacitance approximation may be made, and a uniform temperature may be
assumed to exist in the sphere at any time.
From Eq. 5.8a, achievement of 90% of the maximum possible thermal energy storage corresponds to

Est
= 0.90 = 1 exp ( t / t )
cVi
3
2700 kg / m 0.075m 950 J / kg K
t = Vc / hAs = Dc / 6h = = 427s.
2
6 75 W / m K
t = t ln ( 0.1) = 427s 2.30 = 984s

From Eq. (5.6), the corresponding temperature at any location in the sphere is
( )
T ( 984s ) = Tg,i + Ti Tg,i exp ( 6ht / Dc )

(
T ( 984s ) = 300C 275C exp 6 75 W / m K 984s / 2700 kg / m 0.075m 950 J / kg K
2 3
)
T ( 984s ) = 272.5C

3
If the product of the density and specific heat of copper is (c)Cu 8900 kg/m 400 J/kgK = 3.56
6 3
10 J/m K, is there any advantage to using copper spheres of equivalent diameter in lieu of aluminum
spheres?

Does the time required for a sphere to reach a prescribed state of thermal energy storage change with
increasing distance from the bed inlet? If so, how and why?
Problem 5.15: Heating of coated furnace wall during start-up.

KNOWN: Thickness and properties of furnace wall. Thermal resistance of ceramic coating
on surface of wall exposed to furnace gases. Initial wall temperature.

FIND: (a) Time required for surface of wall to reach a prescribed temperature, (b)
Corresponding value of coating surface temperature.

ASSUMPTIONS: (1) Constant properties, (2) Negligible coating thermal capacitance, (3)
Negligible radiation.
3
PROPERTIES: Carbon steel: = 7850 kg/m , c = 430 J/kgK, k = 60 W/mK.
ANALYSIS: Heat transfer to the wall is determined by the total resistance to heat transfer
from the gas to the surface of the steel, and not simply by the convection resistance.
1 1
1 1
+ 102 m 2 K/W
1
Hence, with U = ( R tot ) = + R f = = 20 W/m 2 K.
h 25 W/m 2 K

UL 20 W/m 2 K 0.01 m
Bi = = = 0.0033 << 1
k 60 W/m K
and the lumped capacitance method can be used.
(a) From Eqs. (5.6) and (5.7),
T T
= exp ( t/ t ) = exp ( t/R t C t ) = exp ( Ut/ Lc )
Ti T

Lc
T T 7850 kg/m3 ( 0.01 m ) 430 J/kg K 1200 1300
t= ln = ln
U Ti T 2
20 W/m K 300 1300

t = 3886s = 1.08h.
(b) Performing an energy balance at the outer surface (s,o),

( ) (
h T Ts,o = Ts,o Ts,i / R f )
hT + Ts,i / R f 25 W/m 2 K 1300 K + 1200 K/10-2 m 2 K/W
Ts,o = =
h + (1/ R f ) ( 25 + 100 ) W/m2 K

Ts,o = 1220 K.

How does the coating affect the thermal time constant?


Transient Conduction:
Spatial Effects and the Role of
Analytical Solutions
Solution to the Heat Equation for a Plane Wall with
Symmetrical Convection Conditions
If the lumped capacitance approximation can not be made, consideration must
be given to spatial, as well as temporal, variations in temperature during the
transient process.
For a plane wall with symmetrical convection
conditions and constant properties, the heat
equation and initial/boundary conditions are:
2T 1 T
=
x 2 t
T ( x,0 ) = T i
T
=0
x x =0

T
k = h T ( L, t ) T
x x=L

Existence of seven independent variables:


T = T ( x, t , T i , T , k , , h )
How may the functional dependence be simplified?
Non-dimensionalization of Heat Equation and Initial/Boundary Conditions:
T T
Dimensionless temperature difference: * =
i Ti T
x
Dimensionless coordinate: x *
L
t
Dimensionless time: t * Fo
L 2

Fo the Fourier Number


hL
The Biot Number: Bi
k solid
* = f ( x * , Fo, Bi )
Exact Solution:

* = C n exp ( n2 Fo ) cos ( n x * )
n =1
4sin n
Cn = n tan n = Bi
2 n + sin ( 2 n )
See Appendix B.3 for first four roots (eigenvalues 1 ,..., 4 ) of Eq. (5.39c)
The One-Term Approximation ( Fo > 0.2 ) :
 Variation of midplane temperature (x*= 0) with time ( Fo ) :

o
*
( To T )
C 1 exp ( 12 Fo )
(T i T )
Table 5.1 C 1 and 1 as a function of Bi
 Variation of temperature with location (x*) and time ( Fo ) :
* = o* cos ( 1 x * )
 Change in thermal energy storage with time:
E st = Q
sin 1 *
Q = Q o 1 o
1
Q o = c (T i T )
Can the foregoing results be used for a plane wall that is well insulated on one
side and convectively heated or cooled on the other?
Can the foregoing results be used if an isothermal condition (T s T i ) is
instantaneously imposed on both surfaces of a plane wall or on one surface of
a wall whose other surface is well insulated?
Placa plana Cilindro longo Esfera

Bi 1 c1 1 c1 1 c1
0.01 0.09983 1.00166 0.14124 1.00250 0.17303 1.00300
0.02 0.14095 1.00331 0.19950 1.00498 0.24446 1.00599
0.03 0.17234 1.00495 0.24403 1.00746 0.29910 1.00898
0.04 0.19868 1.00657 0.28143 1.00993 0.34503 1.01197
0.05 0.22176 1.00819 0.31426 1.01240 0.38537 1.01495
0.06 0.24253 1.00979 0.34383 1.01485 0.42173 1.01793
0.07 0.26153 1.01138 0.37092 1.01729 0.45506 1.02090
0.08 0.27913 1.01297 0.39603 1.01973 0.48600 1.02387
0.09 0.29557 1.01454 0.41954 1.02216 0.51497 1.02684
0.10 0.31105 1.01609 0.44168 1.02458 0.54228 1.02980
0.15 0.37788 1.02372 0.53761 1.03655 0.66086 1.04453
0.20 0.43284 1.03109 0.61697 1.04830 0.75931 1.05915
0.25 0.48009 1.03819 0.68559 1.05984 0.84473 1.07365
0.30 0.52179 1.04505 0.74646 1.07116 0.92079 1.08802
0.35 0.55922 1.05166 0.80140 1.08226 0.98966 1.10226
0.40 0.59324 1.05804 0.85158 1.09314 1.05279 1.11635
0.45 0.62444 1.06419 0.89783 1.10381 1.11118 1.13030
------- ------- ------- ------- ------- ------- -------
Graphical Representation of the One-Term Approximation
The Heisler Charts Plane wall
Midplane Temperature:
Temperature Distribution:

Change in Thermal Energy Storage:


Radial Systems

Long Rods Heated or Cooled by Convection.

Bi = hr o / k
Fo = t / r o2

(5.184a)

Long rod:

(r , t ) T (r , t ) T
=
*
i
=
Ti T
(
= c n J o ( n r *) exp n2 Fo )
n =1

J 1 ( n )
cn =
2 n = n ro r * = r ro
n J o2 ( n ) + J 12 ( n )
Radial Systems

Long rod one term approximation (Fo > 0.2):

* =
T T
Ti T
( )
c1 J o ( 1 r *) exp 12 Fo = o* J o ( 1 r *) o* =
To T
Ti T
(
= c1 exp 12 Fo )

Change in thermal energy storage with time:

E st = Q

(5.184a)

Q 2 J 1 ( 1 ) *
=1 o (Fo > 0.2)
Qo 1

Q o = c (T i T )
Graphical Representation of the One-Term Approximation
The Heisler Charts Infinite cylinder
Centerline Temperature:
Temperature Distribution:

Change in Thermal Energy Storage:


Spherical Systems
Spheres Heated or Cooled by Convection.

Bi = hr o / k
Fo = t / r o2

Sphere: (5.184a)

(r , t ) T (r , t ) T
=
*
i
=
Ti T
(
= c n exp n2 Fo) 1
n r*
( )
sin n r *
n =1

4 (sin n n cos n )
cn =
2 n sin (2 n )
r * = r ro

1 n cotan n = Bi
Spherical Systems

Sphere one term approximation (Fo > 0.2):

sin ( 1 r *) sin ( 1 r *)
* =
T T
Ti T
c1
1 r *
( )
exp 12 Fo = o*
1 r *

o* =
To T
Ti T
(
= c1 exp 12 Fo )
Change in thermal energy storage with time:

E st = Q
(5.184a)

3 o*
= 1 3 (sin 1 1 cos 1 )
Q (Fo > 0.2)
Qo 1

Q o = c (T i T )
Graphical Representation of the One-Term Approximation
The Heisler Charts Sphere
Center Temperature:
Temperature Distribution:

Change in Thermal Energy Storage:


The Semi-Infinite Solid
A solid that is initially of uniform temperature Ti and is assumed to extend
to infinity from a surface at which thermal conditions are altered.
Problem formulation

2 T 1 dT
= T(x, 0) = Ti
x2 dt

T(, t) = Ti

Special Cases:

Case 1: Change in Surface Temperature (Ts)

T ( 0, t ) = T s T ( x,0 ) = T i

T ( x, t ) T s x
= erf
Ti Ts 2 t
k (T s T i )
qs =
t
Case 2: Uniform Heat Flux ( qs = qo )

2qo ( t / )
1
x2 2
T ( x, t ) T i = exp
k 4 t
q x x
o erfc
k 2 t (5.59)

Case 3: Convection Heat Transfer ( h, T )

T
k = h T T ( 0, t )
x x=0

T ( x, t ) T i x
= erfc
T Ti 2 t
hx h 2 t x h t
exp + erfc +

k k k
2
2 t (5.60)
Contact between two semi-infinite bodies
Two bodies initially at uniform temperatures,
TA and TB, are placed in contact at their free
surfaces

If the contact resistance is neglibible, then the


temperature and the heat flux must be equal at
the contact point

k A (Ts T A,i ) k B (Ts TB,i )


=
A t B t

k A A c p, A T A,i + k B B c p, B T B,i
Ts =
k A A c p, A + k B B c p,B
Multidimensional Effects
Solutions for multidimensional transient conduction can often be expressed
as a product of related one-dimensional solutions for a plane wall, P(x,t),
an infinite cylinder, C(r,t), and/or a semi-infinite solid, S(x,t). See Equations
(5.64) to (5.66) and Fig. 5.11.

Consider superposition of solutions for two-dimensional conduction in a


short cylinder:

T ( r , x, t ) T
= P ( x, t ) x C ( r , t )
Ti T
T ( x, t ) T T ( r,t ) T
= x
Ti T Plane Ti T Infinite
Wall Cylinder
T (x,y,t ) T T (x,t ) T T ( y,t ) T
=
Ti T rectangula
barra de seco
r (2 a2b )
Ti T placa infinita de Ti T
espessura 2a
espessura
placa infinita de
2b

T (x,t ) T To (t ) T T ( y,t ) T To (t ) T
=
To (t ) T Ti T placa infinita de To (t ) T
espessura 2 a
Ti T placa infinita de
espessura 2b
Q Q Q
= +
Qo sec o Q
rectangular 2a2b o espessura 2 a
barra de placa plana de Q
o espessura
placa plana de
2b

Q Q

Qo placa plana de Qo
espessura 2a
espessura
placa plana de
2b

T ( x, t ) T T ( x, t ) T T (r , t ) T
S ( x, t ) = P ( x, t ) = C (r , t ) =
Ti T Ti T Ti T
Problem 5.66: Charging a thermal energy storage system consisting of
a packed bed of Pyrex spheres.

KNOWN: Diameter, density, specific heat and thermal conductivity of Pyrex


spheres in packed bed thermal energy storage system. Convection coefficient and
inlet gas temperature.

FIND: Time required for sphere to acquire 90% of maximum possible thermal
energy and the corresponding center and surface temperatures.
SCHEMATIC:
ASSUMPTIONS: (1) One-dimensional radial conduction in sphere, (2)
Negligible heat transfer to or from a sphere by radiation or conduction due to
contact with adjoining spheres, (3) Constant properties.
ANALYSIS: With Bi h(ro/3)/k = 75 W/m2K (0.0125m)/1.4 W/mK = 0.67,
the lumped capacitance method is inappropriate and the approximate (one-term)
solution for one-dimensional transient conduction in a sphere is used to obtain the
desired results.
To obtain the required time, the specified charging requirement
( Q / Q o = 0.9 ) must first be used to obtain the dimensionless center temperature,
o* .

From Eq. (5.52),


13 Q
o = 1
3 sin ( 1 ) 1 cos ( 1 ) Qo

With Bi hro/k = 2.01, 1 2.03 and C1 1.48 from Table 5.1. Hence,
0.1( 2.03)
3
0.837
o = = = 0.155
3 0.896 2.03 ( 0.443) 5.386
From Eq. (5.50c), the corresponding time is
ro2 o
t = 2 ln
1 C1

( )
= k / c = 1.4 W / m K / 2225 kg / m 3 835 J / kg K = 7.54 10 7 m 2 / s,

t=
( 0.0375m ) ln ( 0.155 /1.48 )
2
= 1,020s
7.54 10 m / s ( 2.03)
7 2 2

From the definition of o* , the center temperature is

( )
To = Tg,i + 0.155 Ti Tg,i = 300C 42.7C = 257.3C

The surface temperature at the time of interest may be obtained from Eq. (5.50b)
with r = 1,

o sin ( 1 ) 0.155 0.896


Ts = Tg,i + ( Ti Tg,i ) = 300C 275C = 280.9C
1 2.03

Is use of the one-term approximation appropriate?


Problem: 5.82: Use of radiation heat transfer from high intensity lamps
( qs = 10 4 W/m 2 ) for a prescribed duration (t=30 min) to assess
ability of firewall to meet safety standards corresponding to
maximum allowable temperatures at the heated (front) and
unheated (back) surfaces.
KNOWN: Thickness, initial temperature and thermophysical properties of
concrete firewall. Incident radiant flux and duration of radiant heating.
Maximum allowable surface temperatures at the end of heating.

FIND: If maximum allowable temperatures are exceeded.

SCHEMATIC:
ASSUMPTIONS: (1) One-dimensional conduction in wall, (2) Validity of semi-
infinite medium approximation, (3) Negligible convection and radiative exchange
with the surroundings at the irradiated surface, (4) Negligible heat transfer from
the back surface, (5) Constant properties.

ANALYSIS: The thermal response of the wall is described by Eq. (5.59)

2 q o ( t / )
1/ 2
x 2 q o x x
T ( x, t ) = Ti + exp erfc
k 4 t k 2 t

7
where, = k / c p = 6.92 10 m / s and for
2

t = 30 min = 1800s, 2qo ( t / ) / k = 284.5 K. Hence, at x = 0,


1/ 2

T ( 0,30 min ) = 25C + 284.5C = 309.5C < 325C

( )
At x = 0.25m, x / 4 t = 12.54, q o x / k = 1, 786K, and x / 2 ( t ) = 3.54.
2 1/ 2

Hence,

( )
T ( 0.25m, 30 min ) = 25C + 284.5C 3.58 106 1786C ( ~ 0 ) 25C
Both requirements are met.

Is the assumption of a semi-infinite solid for a plane wall of finite thickness


appropriate under the foregoing conditions?

COMMENTS: The foregoing analysis may or may not be conservative, since


heat transfer at the irradiated surface due to convection and net radiation
exchange with the environment has been neglected. If the emissivity of the
surface and the temperature of the surroundings are assumed to be = 1 and Tsur
= 298K, radiation exchange at Ts = 309.5C would be
(
qrad = Ts4 Tsur
4
)= 6, 080 W / m 2 K,
which is significant (~ 60% of the prescribed radiation). However, under actual
conditions, the wall would likely be exposed to combustion gases and adjoining
walls at elevated temperatures.
5.89
Um cilindro de cobre, com 100 mm de comprimento e 50 mm de dimetro
encontra-se inicialmente temperatura uniforme de 20C.
As duas bases so aquecidas muito rapidamente, a partir de um determinado
instante, ficando temperatura de 500 C, enquanto a superfcie lateral do
cilindro aquecida por uma corrente de gs a 500 C e com um coeficiente de
conveco de 100 W/m2K.

a) Determinar a temperatura do centro do cilindro ao fim de 8 segundos.


b) Atendendo aos parmetros adimensionais que determinam a distribuio de
temperaturas nos problemas de difuso transiente do calor, possvel admitir
hipteses simplificativas na anlise deste problema?
Apresente uma explicao resumida.

Propriedades do cobre
CILINDRO CURTO: 2D
PROPRIEDADES CONSTANTES
h CONSTANTE

PARA O CILINDRO INFINITO C(r,t):

PARA O PLACA PLANA INFINITA P(x,t):

PARA O CILINDRO CURTO:


PARMETROS ADIMENSIONAIS QUE CONTROLAM A CONDUO TRANSIENTE:
Fourier e Biot.

NO CASO DO CILINDRO Bi < 0,1 DESPREZAM-SE GRADIENTES


RADIAIS
5.90
Considerando que a carne fica cozida quando atinge uma temperatura de 80C,
calcule o tempo necessrio para assar uma pea de carne com 2,25 kg.
Admitir que a pea de carne um cilindro com dimetro igual ao comprimento e
que as suas propriedades so equivalentes s de gua lquida.
Considere que a carne se encontra inicialmente temperatura de 6C e que a
temperatura do forno 175C e o coeficiente de conveco de 15 W/m2K.

Propriedades da gua:
CLCULO DAS DIMENSES DO CILINDRO:

CLCULO DA TEMPERATURA NO CENTRO DO CILINDRO:


SOLUO TENTATIVA-ERRO:
Introduction to Convection:
Flow and Thermal Considerations
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
u( y)
and the free stream whose = 0.99
thickness increases in u
the flow direction.
u s
Why does increase in the flow direction? s = Cf =
y =0 1
y u 2
Manifested by a surface shear 2
stress s that provides a drag FD = s dAs
force, FD . As

How does s vary in the flow


direction? Why?
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 Ts T ( y )
t = 0.99
the free stream whose thickness t Ts T
increases in the flow direction.
Why does t increase in the T
qs = k f y =0
flow direction? y
Manifested by a surface heat
flux qs and a convection heat k f T / y y =0
transfer coefficient h . h
Ts T
If (Ts T ) is constant, how do qs and
h vary in the flow direction?
Distinction between Local and
Average Heat Transfer Coefficients
Local Heat Flux and Coefficient:

q = h (Ts T )

Average Heat Flux and Coefficient for a Uniform Surface Temperature:

q = hAs (Ts T )

q = As qdAs = (Ts T ) A hdAs


s

1
h= hdAs
As As

For a flat plate in parallel flow:


1
h = oL hdx
L
Governing equations
Equao da continuidade ( u ) ( v ) ( w)
+ + + =0
t x y z

Equao de balano da quantidade de (


( ui ) u j ui )
p ij
+ = + + gi
movimento t xj xi x j

ui u j 2 uk
ij = +
x j xi 3 x k ij

Equao de conservao da
energia
T u j ui
Energia interna
t xj
(
( e ) + u j e =
xj
) k
xj
p
xj
+ ij
xj
+ q&

ui u u j ui 2 u k 2
Dissipao = ij = i + =
xj x j xi x j 3 x k

viscosa de energia
u 2 v 2 w
2 u v 2 u w 2 v w 2
= 2 + + + + + + + +
x y z y x z x z y

2
2 u v w
+ +
3 x y z
Governing equations
p
h =e+

T p
Entalpia especfica
t xj
(
( h ) + u j h =
xj
) k
xj
+
t
+uj
p
xj
+ + q&

dh = c p dT + (1 T )
1
dp

1
Coeficiente de expanso trmica: =
T p
Gs perfeito: = 1/T

Fluido incompressvel: = 0

T
Temperatura cp

t
(
( T ) + c p u j T =
xj
)
xj
k
xj
+ T p + u j p + + q&
t x j

The Boundary Layer Equations

Consider concurrent velocity and thermal boundary layer development for steady,
( )
two-dimensional, incompressible flow with constant fluid properties , c p , k and
negligible body forces.
Apply conservation of mass, Newtons 2nd Law of Motion and conservation of energy
to a differential control volume and invoke the boundary layer approximations.
Velocity Boundary Layer:
u >> v
u u v v
>> , ,
Thermal Boundary Layer:
y x y x
T T
>>
y x
Conservation of Mass:
u v
+ =0
x y
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?
Newtons Second Law of Motion:
x-direction :
u u dp 2u
u + v = + 2
x u dx y

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

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


y-direction :
p
=0
y
Conservation of Energy:
2
T T 2T u
cp u +v = k +
x y y 2 y

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

What is the second term on the right-hand side called and under what conditions
may it be neglected?
Boundary Layer Similarity
As applied to the boundary layers, the principle of similitude 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:
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: ,
Thermal : c p , k

Hence,
u = f ( x , y , L, V , , )
s = f ( x , L, V , , )
and
T = f ( x , y , L, V , , , c p , k )
h = f ( x , L, V , , , c p , 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 y
x* y*
L L
u v
u* v*
V V
T Ts
T*
T Ts

Neglecting viscous dissipation, the following normalized forms of the x-momentum


* u * u 1 2u *
and energy equations are obtained: * *
dp*
u +v = * +
x *
y *
dx Re L y*2
T * * T
*
1 2T *
u*
+v =
x *
y *
Re L Pr y*2
VL VL
Re L = the Reynolds Number
v
cp v
Pr = the Prandtl Number
k
How may the Reynolds and Prandtl numbers be interpreted physically? Pr n n>0
t
For a prescribed geometry,
(
u * = f x* , y* , Re L )
u V u
*
s = = *
y y =0 L y y* = 0

The dimensionless shear stress, or local friction coefficient, is then


s
2 u *
Cf =
V 2 / 2 Re L y* y* = 0

u *
y*
(
= f x* , Re L )
y* = 0

Cf =
2
Re L
(
f x* , Re L )
What is the functional dependence of the average friction coefficient, Cf ?
For a prescribed geometry,

(
T * = f x* , y* , Re L , Pr )
k f T / y k f (T Ts ) T * k f T *
y =0
h= = =+
Ts T L (Ts T ) y* y* = 0
L y* y* = 0

The dimensionless local convection coefficient is then

hL T *
Nu
kf
= *
y
(
= f x* , Re L , Pr )
y* = 0

Nu local Nusselt number

What is the functional dependence of the average Nusselt number?

How does the Nusselt number differ from the Biot number?
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:
u x
Re x , c c critical Reynolds number

xc location at which transition to turbulence begins
105 < Re x , c < 3 x 106
~ ~
What may be said about transition if ReL < Rex,c? If ReL > Rex,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?
The Reynolds Analogy
Equivalence of dimensionless momentum and energy equations for
negligible pressure gradient (dp*/dx*~0) and Pr~1:

u * * u
*
1 2u *
u*
+v =
x* y* Re y*2

Advection terms Diffusion

T * * T
*
1 2T *
u*
+v =
x *
y *
Re y*2

Hence, for equivalent boundary conditions, the solutions are of the same form:
u* = T *
u * T *
= *
y* y* = 0
y y* = 0

Re
Cf = Nu
2
or, with the Stanton number defined as,
h Nu
St =
Vc p Re Pr

With Pr = 1, the Reynolds analogy, which relates important parameters of the velocity
and thermal boundary layers, is
Cf
= St
2
Modified Reynolds (Chilton-Colburn) Analogy:
An empirical result that extends applicability of the Reynolds analogy:
Cf 2
= St Pr 3
jH 0.6 < Pr < 60
2
Colburn j factor for heat transfer

Applicable to laminar flow if dp*/dx* ~ 0.

Generally applicable to turbulent flow without restriction on dp*/dx*.


Problem 6.28: 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 ( As L ) .

SCHEMATIC:

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,


hL
Nu = = f ( Re L , Pr ) .
k
The Reynolds numbers for the blades are

Re L,1 = ( V1L1 /1 ) = 15m 2 / s 1 ReL,2 = ( V2 L2 / 2 ) = 15m 2 / s 2 .

Hence, with constant properties ( v1 = v2 ) , ReL,1 = ReL,2 . Also, Pr1 = Pr2

Therefore,
Nu 2 = Nu 1
( h 2 L2 / k 2 ) = ( h1L1 / k1 ) L L q1
h 2 = 1 h1 = 1
L2 (
L 2 A1 Ts,1 T )

The heat rate for the second blade is then


L A 2 Ts,2 T ( )
(
q 2 = h 2 A 2 Ts,2 T = 1 ) q1
L 2 A1 Ts,1 T ( )
Ts,2 T ( 400 35)
q2 = q1 = (1500 W )
Ts,1 T ( 300 35) q 2 = 2066 W.

COMMENTS: (i) The variation in from Case 1 to Case 2 would cause ReL,2 to differ from
ReL,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 h2.
Problem 6.35: 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, Ts.

SCHEMATIC:
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.

PROPERTIES: Table A-4, Air (Evaluate properties at the average temperature of air in the boundary
layer. Assuming Ts = 45C, Tave = (45 + 25)/2 = 35C = 308K. Also, p = 1atm): = 16.69
-6 2 -3
10 m /s, k = 26.9 10 W/mK, Pr = 0.703.

ANALYSIS: From an energy balance on the chip,


q conv = E& g = 30mW.

Newtons law of cooling for the upper chip surface can be written as
Ts = T + q conv / h A chip (2)
2
where A chip = l .

( )
Assuming that the average heat transfer coefficient h over the chip surface is equivalent to the local
coefficient evaluated at x = L, that is, h chip h x ( L ) , the local coefficient can be evaluated by
applying the prescribed correlation at x = L.
0.85
hx x Vx
Nu x = = 0.04 Pr1/ 3
k
0.85
k VL
h L = 0.04 Pr1/ 3
L
0.85
0.0269 W/m K 10 m/s 0.120 m
h L = 0.04 ( 0.703)1/ 3 = 107 W/m2 K.
0.120 m 16.69 10-6 m 2 / s

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


Ts = 25o C + 30 10-3 W/107 W/m 2 K ( 0.004m ) = 42.5o C.
2

COMMENTS: (1) The estimated value of Tave used to evaluate the air properties is reasonable.
(2) How else could h chip have been evaluated? Is the assumption of h = hL reasonable?
External Flow:
The Flat Plate in Parallel Flow
Physical Features

As with all external flows, the boundary layers develop freely without constraint.
Boundary layer conditions may be entirely laminar, laminar and turbulent,
or entirely turbulent.
To determine the conditions, compute
u L u L
Re L = =

and compare with the critical Reynolds number for transition to turbulence, Re x , c .
Re L < Re x , c laminar flow throughout

Re L > Re x , c transition to turbulent flow at xc / L Re x , c / Re L


Value of Re x , c depends on free stream turbulence and surface roughness.
Nominally,
Re x , c 5 105.

If boundary layer is tripped at the leading edge


Re x , c = 0
and the flow is turbulent throughout.

Surface thermal conditions are commonly idealized as being of uniform


temperature Ts or uniform heat flux qs . Is it possible for a surface to be
concurrently characterized by uniform temperature and uniform heat flux?
Thermal boundary layer development may be delayed by an unheated
starting length.

Equivalent surface and free stream temperatures for x < and uniform Ts
(or qs ) for x > .
Similarity Solution for Laminar,
Constant-Property Flow over an Isothermal Plate
Based on premise that the dimensionless x-velocity component, u / u ,
and temperature, T * (T Ts ) / (T Ts ) , can be represented exclusively in
terms of a dimensionless similarity parameter
y ( u / x )
1/ 2

Similarity permits transformation of the partial differential equations associated


with the transfer of x-momentum and thermal energy to ordinary differential
equations of the form
d3 f d2 f
2 3+ f =0
d d 2

where ( u / u ) df / d , and

d 2T * Pr dT *
+ f =0
d 2
2 d
Similarity Solution for Laminar,
Constant-Property Flow over an Isothermal Plate
Subject to prescribed boundary conditions, numerical solutions to the momentum
and energy equations yield the following results for important local boundary layer
parameters:
- with u / u = 0.99 at = 5.0,
5.0 5x
= =
( u / vx )
1/ 2
Re x
u d2 f
- with s = = u u / vx
y y =0
d 2 =0

and d 2 f / d 2 = 0.332,
=0

s, x
C f ,x = 0.664Re x1/ 2
u2 / 2
- with hx = qs / (Ts T ) = k T * / y = k ( u / vx ) dT * / d
1/ 2
y =0 =0

and dT * / d = 0.332 Pr1/ 3 for Pr > 0.6,


=0

hx
Nu x = x = 0.332 Re1/x 2 Pr1/ 3
and k

= Pr1/ 3
t
How would you characterize relative laminar velocity and thermal boundary layer
growth for a gas? An oil? A liquid metal?

How do the local shear stress and convection coefficient vary with distance from
the leading edge?
Average Boundary Layer Parameters:
1
s , x 0x s dx
x
Cf , x = 1.328 Re x1/ 2

1 x
hx = 0 hx dx
x
Nu x = 0.664 Re1/x 2 Pr1/ 3
The effect of variable properties may be considered by evaluating all properties
at the film temperature.
Ts + T
Tf =
2
Turbulent Flow
Local Parameters:

Empirical
C f , x = 0.0592 Rex1/ 5
Correlations
Nu x = 0.0296 Re 4x / 5 Pr1/ 3

How do variations of the local shear stress and convection coefficient with
distance from the leading edge for turbulent flow differ from those for laminar flow?

Average Parameters:
hL =
1 xc
L
(
0 h1am dx + xc hturb dx
L
)
Substituting expressions for the local coefficients and assuming Re x ,c = 5 105 ,
0.074 1742
C f , L = 1/ 5
Re L Re L
( )
Nu L = 0.037 Re4L / 5 871 Pr1/ 3
For Re x , c = 0 or L xc ( Re L Re x , c ) ,
C f , L = 0.074 Re L1/ 5
Nu L = 0.037 Re 4L / 5 Pr1/ 3
Special Cases: Unheated Starting Length (USL)
and/or Uniform Heat Flux

For both uniform surface temperature (UST) and uniform surface heat flux (USF),
the effect of the USL on the local Nusselt number may be represented as follows:
Laminar Turbulent
Nu x = 0 UST USF UST USF
Nu x = a 3/4 3/4 9/10 9/10
a b
1 ( / x )
b 1/3 1/3 1/9 1/9
C 0.332 0.453 0.0296 0.0308
Nu x = 0 = C Re mx Pr1/ 3
m 1/2 1/2 4/5 4/5

Sketch the variation of hx versus ( x ) for two conditions: > 0 and = 0.


What effect does an USL have on the local convection coefficient?
UST:
qs = hx (Ts T ) q = hL As (Ts T )
L ( 2 p +1) / ( 2 p + 2 ) 2 p / ( 2 p +1)
Nu L = Nu L 1 ( / L )
=0 ( L )
p = 1 for laminar flow throughout
p = 4 for turbulent flow throughout

hL numerical integration for laminar/turbulent flow


1 xc
hL = h1am dx + xL hturb dx
L c
USF:
qs
Ts = T + q = qs As
hx

Treatment of Non-Constant Property Effects:


Evaluate properties at the film temperature.
T + T
Tf = s
2
Problem 7.21: Preferred orientation (corresponding to lower heat loss) and the
corresponding heat rate for a surface with adjoining smooth and
roughened sections.

SCHEMATIC:

ASSUMPTIONS: (1) Surface B is sufficiently rough to trip the boundary layer when in the upstream position
(Configuration 2); (2) Re x, c 5 105 for flow over A in Configuration 1.
-6 2 -3
PROPERTIES: Table A-4, Air (Tf = 333K, 1 atm): = 19.2 10 m /s, k = 28.7 10
W/mK, Pr = 0.7.
ANALYSIS: Since Configuration (2) results in a turbulent boundary layer over the entire
surface, the lowest heat transfer is associated with Configuration (1).
Find
u L 20 m/s 1m
Re L = = = 1.04 106.
19.2 10-6 m 2 / s

Hence in Configuration (1), transition will occur just before the rough surface (xc = 0.48m).


( )
4/5
Nu L,1 = 0.037 1.04 106 871 0.71/3 = 1366

h L,1L
For Configuration (1): = Nu L,1 = 1366.
k

Hence
( )
h L,1 = 1366 28.7 103 W/m K /1m = 39.2 W/m 2 K

q1 = h L,1A ( Ts T ) = 39.2 W/m 2 K ( 0.5m 1m )(100 20 ) K = 1568 W <

( )
4/5
Comment: Note that Nu L,2 = 0.037 1.04 106 ( 0.7 )1/ 3 = 2139 > Nu L,1 .
External Flow:
Flow over Bluff Objects
(Cylinders, Sphere)
The Cylinder in Cross Flow
Conditions depend on special features of boundary layer development, including
onset at a stagnation point and separation, as well as transition to turbulence.

Stagnation point: Location of zero velocity ( u = 0 ) and maximum pressure.

Followed by boundary layer development under a favorable pressure gradient


( dp / dx < 0 ) and hence acceleration of the free stream flow ( du / dx > 0 ) .
As the rear of the cylinder is approached, the pressure must begin to increase.
Hence, there is a minimum in the pressure distribution, p(x), after which boundary
layer development occurs under the influence of an adverse pressure gradient
( dp / dx > 0, du / dx < 0 ) .
Separation occurs when the velocity gradient du / dy y =0 reduces to zero.

and is accompanied by flow reversal and a downstream wake.

Location of separation depends on boundary layer transition.

VD VD
Re D =

What features differentiate boundary development for the flat plate in
parallel flow from that for flow over a cylinder?

Force imposed by the flow is due to the combination of friction and form drag.
FD form of the drag force is
The dimensionless
CD = Figure 7.8
(
Af V / 2
2
)
Heat Transfer Considerations
Heat Transfer Considerations
The Local Nusselt Number:
How does the local Nusselt number vary with for Re D < 2 x 105 ?
What conditions are associated with maxima and minima in the variation?
How does the local Nusselt vary with for ReD > 2 x 105 ? What conditions
are associated with maxima and minima in the variation?

(
The Average Nusselt Number Nu D hD / k : )
Churchill and Bernstein Correlation:
4/5
0.62 Re 1/ 2
Pr 1/ 3 Re
5/8

Nu D = 0.3 + D
1/ 4 1 + D

1 + ( 0.4 / Pr )2 / 3 282,000

ReD C m
Cylinders of Noncircular Cross Section: 0.4 4 0.989 0.330
4 40 0.911 0.385
Nu D = C Re mD Pr1/ 3
C , m Table 7.3 40 4103 0.683 0.466
4103 4104 0.193 0.618
4104 4105 0.027 0.805
Flow Across Tube Banks
A common geometry for Aligned and Staggered Arrays:
two-fluid heat exchangers.

ST
Aligned: Vmax = V
ST D

( ) ( )
Staggered: ST
Vmax = V if 2 S D D ST D
ST D

( ) ( )
ST
or, Vmax = V if 2 S D D ST D
(
2 SD D)
Flow Conditions:

How do convection coefficients vary from row-to-row in an array?


How do flow conditions differ between the two configurations?
Why should an aligned array not be use for ST/SL < 0.7?
Average Nusselt Number for an Isothermal Array:

Nu D = C2 C Re mD ,max Pr 0.36 ( Pr/ Prs )


1/ 4

C , m Table 7.7
C2 Table 7.8
All properties are evaluated at (Ti + To ) / 2 except for Prs.
Geometria ReD,max C m
10 - 102 0.80 0.40
Aproximado por um tubo
102 103
isolado
Tubos alinhados
103 - 2105 (ST /SL <
0.27 0.63
0.7)*
2105 - 2106 0.021 0.84
10 - 102 0.90 0.40
Aproximado por um tubo
102 103
isolado
Tubos
103 - 2105 (ST / SL < 2) 0.35 (ST / SL)1/5 0.60
desfasados

103 - 2105 (ST / SL > 2) 0.40 0.60

2105 - 2106 0.022 0.84

NL 1 2 3 4 5 7 10 13 16

Alinhados 0.70 0.80 0.86 0.90 0.92 0.95 0.97 0.98 0.99

Desfasados 0.64 0.76 0.84 0.89 0.92 0.95 0.97 0.98 0.99
Fluid Outlet Temperature (To) :
Ts To DNh
= exp
Ts Ti VNT ST c p

N = NT x N L

What may be said about To as N ?

Total Heat Rate:


q = hAs Tlm
As = N ( DL )
(T T ) (Ts To )
Tlm = s i
Ts Ti
ln
Ts To
Pressure Drop:
Vmax
2

p = N L f
2
, f Figures 7.13 and 7.14
The Sphere
Flow over a sphere
Boundary layer development is similar to that for flow over a cylinder,
involving transition and separation.

( )
Nu D = 2 + 0.4 Re1/D 2 + 0.06 Re2D/ 3 Pr 0.4 ( / s )
1/ 4

CD Figure 7.8
Problem: 7.78 Measurement of combustion gas temperature with a spherical
thermocouple junction.

KNOWN: Velocity and temperature of combustion gases. Diameter and emissivity of thermocouple
junction. Combustor temperature.

FIND: (a) Time to achieve 98% of maximum thermocouple temperature rise for negligible radiation, (b)
Steady-state thermocouple temperature, (c) Effect of gas velocity and thermocouple emissivity on
measurement error.

SCHEMATIC:
ASSUMPTIONS: (1) Validity of lumped capacitance analysis, (2) Constant properties, (3) Negligible
conduction through lead wires, (4) Radiation exchange between small surface and a large enclosure (parts b
and c).

PROPERTIES: Thermocouple: 0.1 1.0, k = 100 W/mK, c = 385 J/kgK, = 8920 kg/m3; Gases:
k = 0.05 W/mK, = 50 10-6 m2/s, Pr = 0.69.

ANALYSIS: (a) If the lumped capacitance analysis may be used, it follows from Equation 5.5 that

Vc T T D c
t= ln i = ln ( 50 ) .
hAs T T 6h

Neglecting the viscosity ratio correlation for variable property effects, use of V = 5 m/s with the Whitaker
correlation yields

(
Nu D = ( hD k ) = 2 + 0.4 Re1D/ 2 + 0.06 Re D
2/3
)
Pr 0.4

h=
0.05 W m K
0.001m (
2 + 0.4 (100 )
1/ 2
+ 0.06 (100 )
2/3
)
( 0.69 ) = 328 W m 2 K
0.4

Since Bi = h ( ro 3) k = 5.5 10-4, the lumped capacitance method may be used.

t=
( )
0.001m 8920 kg m3 385 J kg K
ln ( 50 ) = 6.83s
6 328 W m 2 K
(b) Performing an energy balance on the junction, qconv = qrad.

Hence, evaluating radiation exchange from Equation 1.7 and with = 0.5,

hAs ( T T ) = As T 4 Tc4 ( )
0.5 5.67 108 W m 2 K 4 4
(1000 T ) K = T ( 400 ) K 4
4
328 W m 2 K

T = 936 K

Parametric calculations to determine the effects of V and yield the following results:

1000 990

970
Temperature, T(K)

Temperature, T(K)
950
950
930

910

900 890
0 5 10 15 20 25 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
Velocity, V(m/s) Emissivity
Emissivity, epsilon = 0.5 Velocity, V = 5 m/s
Since the temperature recorded by the thermocouple junction increases with increasing V and decreasing ,
the measurement error, T - T, decreases with increasing V and decreasing . The error is due to net
radiative transfer from the junction (which depresses T) and hence should decrease with decreasing .

For a prescribed heat loss, the temperature difference ( T - T) decreases with decreasing convection
resistance, and hence with increasing h(V).

COMMENTS: To infer the actual gas temperature (1000 K) from the measured result (936 K),
correction would have to be made for radiation exchange with the cold surroundings.

What measures may be taken to reduce the error associated with radiation effects?
Internal Flow:
General Considerations
Entrance Conditions
Must distinguish between entrance and fully developed regions.
Hydrodynamic Effects: Assume laminar flow with uniform velocity profile at
inlet of a circular tube.

Velocity boundary layer develops on surface of tube and thickens with increasing x.
Inviscid region of uniform velocity shrinks as boundary layer grows.
 Does the centerline velocity change with increasing x? If so, how does it change?
Subsequent to boundary layer merger at the centerline, the velocity profile
becomes parabolic and invariant with x. The flow is then said to be
hydrodynamically fully developed.
 How would the fully developed velocity profile differ for turbulent flow?
Thermal Effects: Assume laminar flow with uniform temperature, T ( r ,0 ) = Ti , at
inlet of circular tube with uniform surface temperature, Ts Ti , or heat flux, qs .

Thermal boundary layer develops on surface of tube and thickens with increasing x.
Isothermal core shrinks as boundary layer grows.

Subsequent to boundary layer merger, dimensionless forms of the temperature


profile ( for Ts and qs ) become independent of x.

 Is the temperature profile invariant with x in the fully developed region?


 For uniform surface temperature, what may be said about the change
in the temperature profile with increasing x?

 For uniform surface heat flux, what may be said about the change in the
temperature profile with increasing x?

 How do temperature profiles differ for laminar and turbulent flow?


The Mean Velocity and Temperature
Absence of well-defined free stream conditions, as in external flow, and hence a
reference velocity ( u ) or temperature (T ) , dictates the use of a cross-
sectional mean velocity ( um ) and temperature (Tm ) for internal flow.

Linkage of mean velocity to mass flow rate:

m = um Ac
or,

m = Ac u ( r , x ) d Ac
Hence,
Ac u ( r , x ) d Ac
um =
Ac

For incompressible flow in a circular tube of radius ro ,

ro
2
um =
2
o u ( r , x ) r dr
r
o
Linkage of mean temperature to thermal energy transport associated with flow
through a cross section:

E t = Ac ucT dAc mcTm


Hence,
Ac ucT dAc
Tm =
m c

For incompressible, constant-property flow in a circular tube,

ro
2
Tm =
um ro2
u ( x, r )T ( x, r ) r dr
0

Newtons Law of Cooling for the Local Heat Flux:


qs = h (Ts Tm )

What is the essential difference between use of Tm for internal flow and T
for external flow?
Hydrodynamic and Thermal Entry Lengths
Entry lengths depend on whether the flow is laminar or turbulent, which, in turn,
depend on Reynolds number.

um Dh
Re D

The hydraulic diameter is defined as
4 Ac
Dh
P
in which case,
um Dh 4 m
Re D =
P

For a circular tube,


um D 4 m
Re D = =
D
Onset of turbulence occurs at a critical Reynolds number of
Re D , c 2300

Fully turbulent conditions exist for


Re D 10,000

Hydrodynamic Entry Length


Laminar Flow: ( x fd , h / D ) 0.05Re D
Turbulent Flow: 10 < ( x fd , h / D ) < 60

Thermal Entry Length


Laminar Flow: ( x fd ,t / D ) 0.05 Re D Pr
Turbulent Flow: 10 < ( x fd ,t / D ) < 60

For laminar flow, how do hydrodynamic and thermal entry lengths compare for a gas?
An oil? A liquid metal?
Fully Developed Conditions
Assuming steady flow and constant properties, hydrodynamic conditions, including
the velocity profile, are invariant in the fully developed region.

What may be said about the variation of the mean velocity with distance from the
tube entrance for steady, constant property flow?

The pressure drop may be determined from knowledge of the friction factor
f, where,
( dp / dx ) D
f
um2 / 2

Laminar flow in a circular tube:


64
f =
Re D

Turbulent flow in a smooth circular tube:


f = ( 0.790 1n Re D 1.64 )
2
Turbulent flow in a roughened circular tube:

Pressure drop for fully developed flow from x1 to x2:


um2
p = p1 p2 = f ( x2 x1 )
2D
and power requirement
pm
P = p =

Requirement for fully developed thermal conditions:
Ts ( x ) T ( r , x )
=0
x Ts ( x ) Tm ( x ) fd ,t

Effect on the local convection coefficient:


Ts T T / r r = r
= o
f ( x)
r Ts Tm r = r Ts Tm
o

Hence, assuming constant properties,


qs / k h
= f ( x)
Ts Tm k

h f ( x)
Variation of h in entrance and fully developed regions:
Determination of the Mean Temperature
Determination of Tm ( x ) is an essential feature of an internal flow analysis.
Determination begins with an energy balance for a differential control volume.

dqconv = m d ( cTm + p ) mc p dTm

Why is the second equality in the foregoing expression considered to be approximate?


Integrating from the tube inlet to outlet,

qconv = m c p (Tm , o Tm ,i ) (1)


A differential equation from which Tm ( x ) may be determined is obtained by
substituting for dq conv = qs ( P dx ) = h (Ts Tm ) P dx.

dTm qs P P
= = h (Ts Tm ) ( 2)
dx mc mcp
p

Special Case: Uniform Surface Heat Flux


dTm qs P
= f ( x)
dx m c
p

qs P
Tm ( x ) = Tm ,i + x
m cp

Why does the surface temperature vary with x as shown in the figure?
In principle, what value does Ts assume at x=0?
Total heat rate:
qconv = qs PL
Special Case: Uniform Surface Temperature
From Eq. (2), with T Ts Tm
d Tm d ( T ) P
= = h T
dx dx m cp
Integrating from x=0 to any downstream location,

Ts Tm ( x ) Px
= exp hx
Ts Tm ,i
mcp
1 x
hx = o hx dx
x
Overall Conditions:

To Ts Tm , o PL hA
= = exp h = exp s
Ti Ts Tm ,i
mcp mcp
qconv = h As Tlm
To Ti
Tlm = ( 3)
1n ( To / Ti )
Special Case: Uniform External Fluid Temperature


To T Tm , o U As = exp 1
= = exp
Ti T Tm ,i
m cp m c p Rtot
T
q = UAs Tlm = lm
Rtot

Tlm Eq. (3) with Ts replaced by T .

Note: Replacement of T by Ts,o if outer surface temperature is uniform.


Problem 8.17: Estimate temperature of water emerging from a thin-walled
tube heated by walls and air of a furnace. Inner and outer
convection coefficients are known.

KNOWN: Water at prescribed temperature and flow rate enters a 0.25 m diameter, black thin-walled
tube of 8-m length, which passes through a large furnace whose walls and air are at a temperature of
Tfur = T = 700 K. The convection coefficients for the internal water flow and external furnace air are
300 W/m2K and 50 W/m2K, respectively.

FIND: The outlet temperature of the water, Tm,o.


SCHEMATIC:

ASSUMPTIONS: (1) Steady-state conditions; (2) Tube is small object with large, isothermal surroundings; (3)
Furnace air and walls are at the same temperature; and (3) Tube is thin-walled with black surface.

PROPERTIES: Table A-6, Water: cp 4180 J/kgK.

ANALYSIS: The linearized radiation coefficient may be estimated from Eq. 1.9 with = 1,

(
h rad ( Tt + Tfur ) Tt2 + Tfur
2
)
where Tt represents the average tube wall surface temperature, which can be estimated from an energy balance
on the tube.

As represented by the thermal circuit, the energy balance may be expressed as


Tm Tt Tt Tfur
=
R cv,i 1/ R cv,o + 1/ R rad

The thermal resistances, with As = PL = DL, are


R cv,i = 1/ h i As R cv,o = 1/ h o As R rad = 1/ h rad
and the mean temperature of the water is approximated as
(
Tm = Tm,i + Tm,o / 2 )
The outlet temperature can be calculated from Eq. 8.46b, with Tfur = T,

T Tm,o
= exp 1
T Tm,i
m cp R tot

where
1
R tot = R cv,i +
1/ R cv,o + 1/ R rad

with
R cv,i = 6.631 105 K / W R cv,o = 3.978 104 K / W R rad = 4.724 104 K / W

it follows that
Tm = 331 K Tt = 418 K Tm,o = 362 K
Internal Flow:
Heat Transfer Correlations
Fully Developed Flow
Laminar Flow in a Circular Tube:
The local Nusselt number is a constant throughout the fully developed
region, but its value depends on the surface thermal condition.
Uniform Surface Heat Flux (qs ) :
Nu D = hD = 4.36
k
Uniform Surface Temperature (Ts ) :
Nu D = hD = 3.66
k
Turbulent Flow in a Circular Tube:
For a smooth surface and fully turbulent conditions ( Re D > 10,000 ) , the
Dittus Boelter equation may be used as a first approximation:
Nu D = 0.023Re 4D/ 5 Pr n n = 0.3 (Ts < Tm )
n = 0.4 (Ts > Tm )
The effects of wall roughness and transitional flow conditions ( Re D > 3000 )
may be considered by using the Gnielinski correlation:
Nu D =
( f / 8 )( Re D 1000 ) Pr
1 + 12.7 ( f / 8 ) ( Pr 2 / 3 1)
1/ 2
Smooth surface:
f = ( 0.790 1n Re D 1.64 )
2

Surface of roughness e > 0 :


f Figure 8.3
Noncircular Tubes:
Use of hydraulic diameter as characteristic length:
4A
Dh c
P
Since the local convection coefficient varies around the periphery of a tube,
approaching zero at its corners, correlations for the fully developed region
are associated with convection coefficients averaged over the periphery
of the tube.
Laminar Flow:
The local Nusselt number is a constant whose value (Table 8.1) depends on
the surface thermal condition (Ts or qs ) and the duct aspect ratio.
Turbulent Flow:
As a first approximation, the Dittus-Boelter or Gnielinski correlation may be used
with the hydraulic diameter, irrespective of the surface thermal condition.
Effect of the Entry Region
The manner in which the Nusselt decays from inlet to fully developed conditions
for laminar flow depends on the nature of thermal and velocity boundary layer
development in the entry region, as well as the surface thermal condition.

Laminar flow in a
circular tube.

Combined Entry Length:


Thermal and velocity boundary layers develop concurrently from uniform
profiles at the inlet.
Thermal Entry Length:
Velocity profile is fully developed at the inlet, and boundary layer development
in the entry region is restricted to thermal effects. Such a condition may also
be assumed to be a good approximation for a uniform inlet velocity profile if
Pr >> 1. Why?

Average Nusselt Number for Laminar Flow in a Circular Tube with Uniform
Surface Temperature:
Combined Entry Length:
Re D Pr/ ( L / D ) ( / s )
1/ 3
> 2:
0.14

0.14

1/ 3
Re Pr
Nu D = 1.86 D
L/ D s
Re D Pr/ ( L / D ) ( / s )
1/ 3
< 2:
0.14

Nu D = 3.66
Thermal Entry Length:
0.0668 ( D / L ) Re D Pr
Nu D = 3.66 +
1 + 0.04 ( D / L ) Re D Pr
2/3
Average Nusselt Number for Turbulent Flow in a Circular Tube :
Effects of entry and surface thermal conditions are less pronounced for
turbulent flow and can be neglected.
For long tubes ( L / D > 60 ) :
Nu D Nu D , fd

For short tubes ( L / D < 60 ) :


Nu D 1 + C
( L / D)
m
Nu D , fd

C 1
m 2/3

Noncircular Tubes:
Laminar Flow:
Nu Dh depends strongly on aspect ratio, as well as entry region and surface
thermal conditions.
Turbulent Flow:
As a first approximation, correlations for a circular tube may be used
with D replaced by Dh .

When determining Nu D for any tube geometry or flow condition, all


properties are to be evaluated at
T m (Tm ,i + Tm , o ) / 2

Why do solutions to internal flow problems often require iteration?


The Concentric Tube Annulus
Fluid flow through
region formed by
concentric tubes.

Convection heat transfer


may be from or to inner
surface of outer tube and
outer surface of inner tube.

Surface thermal conditions may be characterized by


uniform temperature (Ts ,i , Ts , o ) or uniform heat flux ( qi, qo ) .

Convection coefficients are associated with each surface, where


qi = hi (Ts ,i Tm )

qo = ho (Ts , o Tm )
hi Dh ho Dh
Nui Nuo
k k
Dh = Do Di

Fully Developed Laminar Flow


Nusselt numbers depend on Di / Do and surface thermal conditions (Tables 8.2, 8.3)

Fully Developed Turbulent Flow


Correlations for a circular tube may be used with D replaced by Dh .
Conveco Natural
Consideraes Gerais
A conveco natural tem lugar quando h movimento de um fluido
resultante de foras de impulso.

A impulso tem lugar num fluido onde h gradientes de densidade e uma


fora mssica (por exemplo, fora gravtica) proporcional densidade.
Em transmisso de calor, os gradientes de densidade so devidos a
gradientes de temperatura e a fora mssica a fora gravtica.

Gradientes de temperatura estveis e instveis


Escoamentos sem superfcie adjacente (esteira, jacto, camada de mistura)
 Ocorre num meio (em princpio, infinito), em repouso (velocidade
nula longe da origem do escoamento).

 Plumas e jactos com impulso:

Escoamentos com superfcie adjacente (camada limite)

 Escoamento de camada limite numa superfcie quente ou fria (Ts T )


induzido por foras de impulso.
Placas verticais
Desenvolvimento da camada limite numa placa vertical aquecida

 Escoamento ascendente com velocidade mxima dentro da camada limite e


velocidade nula na superfcie da placa e na extremidade (y = ).

 Quais as diferenas relativamente a conveco forada?

 Quais as diferenas relativamente a uma placa arrefecida (Ts < T) ?


Equao de balano de quantidade de movimento na direco x
(escoamento laminar)

u u 1 p 2 u
u +v = g +
x y x y2

p p
=
x dentro da camada limite x fora da camada limite

p 1 1
= g =
x T p T T

u u g 2 u
u +v = ( ) +
x y y2

u u 2 u
u +v = g (T T ) +
x y y2
Equao de balano de quantidade de movimento na direco x
(escoamento laminar)
u
u + u 2
= g (T T ) + u2
x y y

Foras de inrcia Fora de impulso Fora viscosa

 Dado que u (x,y) depende de T (x,y), a soluo desta equao tem de ser
obtida juntamente com a soluo para a equao de camada limite da
energia T (x,y).

u T + T 2
= T2
x y y

As solues esto acopladas.


Adimensionalizao das equaes
x y
x* = y* =
L L
u v T T
u* = v* = T* =
uo uo Ts T

u* u * g (Ts T ) L 1 2 u *
u* + v* = T *+
x* y* uo2 Re L y *2

T * T * 1 2 T *
u* + v* =
x* y * Re L Pr y *2

g (Ts T ) L uo L g (Ts T ) L3
2

GrL = =
uo2 1
424 3 2

Re 2

u* u * GrL 1 2 u *
u* + v* = 2 T *+
x* y * Re Re L y *2
Parmetros adimensionais relevantes

 Nmero de Grashof:
g (Ts T ) L3 Foras de impulso
GrL = ~
2
Foras viscosas
L: dimenso caracterstica da superfcie

= 1
Coeficiente de expanso trmica da superfcie (propriedade
T p termodinmica do fluido

Lquidos:  Tabelas A.5, A.6 de Incropera e de Witt

Gs perfeito: = 1/T (K)

 Rayleigh Number

g (Ts T ) L3
RaL = GrL Pr =

Mtodo integral
Equao de balano integral de quantidade de movimento:

d u
0 u dy = 0 g (T T ) dy
2

dx y y =0

Equao de balano integral de energia:



(T T ) u dy = T


d
d x 0 y y =0

Exemplo de aplicao: placa vertical isotrmica

Vamos assumir um perfil de velocidades cbico

u ( x,0) = 0 u ( x, ) = 0

u 2 u
= 0 2 = g (Ts T )
y y = y y =0
e um perfil de temperaturas quadrtico
T(x,0)=Ts T
= 0
y y =
T(x,)=T

Perfis de velocidades e de temperatura

g (Ts T ) 2 y y
2 2
y y
u= 1 = u o ( x ) 1
4

T T y
2

= 1
Ts T

Substituindo nas equaes de balano integral e integrando resulta


u
1 d 2
105 d x
( )
uo = g (Ts T ) o
1
3
2
1 d
30 d x
( )
uo =

Vamos assumir que uo e so funes do tipo

uo ( x ) = C1 x m ( x ) = C2 x n

daqui resulta
2m + n 2
C1 C2 x 2 m + n 1 = 2 g (Ts T ) x n 1 x m n
C C
105 3 C2
m+n 2 n
C1 C2 x m + n 1 = x
30 C2

Para as equaes estarem dimensionalmente correctas, o expoente de x tem de


ser o mesmo em todos os termos de cada equao, de onde resulta

2m + n 1 = n = m n
m=1/2, n=1/4
m + n 1 = n
g (Ts T )
1 2
Logo 20
12

C1 = 5.17 + Pr
21 2
g (Ts T )
14 1 4
20
C2 = 3.93 + Pr Pr 1 2
21 2
Obtm-se ento
= 5.17 (0.952 + Pr ) Grx
uo 1 2 12


= 3.93 Pr 1 2 (0.952 + Pr ) Grx
14 1 4

pelo que
T
k
y y =0 x 2 x
= 0.508 Pr1 2 (0.952 + Pr ) Grx
hx 1 4
Nu x = = =
14

k Ts T k

ou, de modo equivalente,


14
Pr
Nu x = 0.508 Ra x
14

0.952 + Pr
Por sua vez,
14
hL Pr
Nu L = = 0.68 Ra L
14

k 0. 952 + Pr
Esta soluo est em bom acordo com a soluo exacta e com dados
experimentais
Soluo de semelhana
 Usando a seguinte varivel de semelhana, a equao de balano
de quantidade de movimento na direco x pode ser transformada de
uma equao com derivadas parciais (em x e y) numa equao
diferencial ordinria expressa exclusivamente em termos de .
1/ 4
x
y Gr
x 4

 Equaes de balano de quantidade de movimento e energia

f + 3 ff 2 ( f ) + T = 0
2

T + 3Pr fT = 0
T T
= x ( Grx1/ 2 ) u
df
f ( ) T
d 2 Ts T
 A integrao numrica das equaes conduz aos seguintes resultados
para f () e T*:

 Espessura da camada limite hidrodinmica ( ) 5 for Pr > 0.6


1/ 4
Gr x
 Pr > 0.6 : = 5 x x = 7.07 x1/ 4
( Grx )
1/ 4
4
 Nmeros de Nusselt ( Nu x and Nu L ) :
1/ 4 1/ 4
Gr
Nu x = hx = x dT Gr
= x g ( Pr )
k 4 d 4
=0

0.75 Pr1/ 2
g ( Pr ) = ( 0 < Pr < )
( 0.609 + 1.221 Pr + 1.238 Pr )
1/ 2 1/ 4

h = 1 oL h dx Nu L = 4 Nu L
L 3

Transio para regime turbulento


 A ampliao de perturbaes
depende do valor relativo das foras
de impulso e das foras viscosas
 A transio ocorre para o seguinte
nmero de Rayleigh crtico:
g (Ts T ) x3
Rax , c = Grx , c Pr = 109

Correlaes empricas (Churchill e Chu)
 Escoamento laminar ( Ra L < 109 ) :

0.670 Ra1/L 4
Nu L = 0.68 + 4/9
1 + ( 0.492 / Pr )9 /16

 Todas as condies
2

0.387 RaL 1/ 6

Nu L = 0.825 +
9 /16 4 / 9
1 + ( 0.492 / Pr )

Placas inclinadas
Componente da acelerao gravtica paralela placa: g cos

T s < T T s > T

Quando o fluido se mantm junto parede, as correlaes de Churchill e


Chu podem ser usadas, desde que 0 60 e substituindo g por g cos

Quando o fluido tem tendncia a afastar-se da parede, o coeficiente de


conveco aumenta e as correlaes apresentadas no so vlidas
Placas Horizontais
A fora de impulso normal s placas

O escoamento e a transmisso de calor dependem de a placa estar


aquecida ou arrefecida e de a troca de calor se dar na face superior ou
inferior.
Face superior de placa aquecida ou Face inferior de placa arrefecida

Ts > T Ts < T
Nu L = 0.54 Ra1/L 4 (10 4
< RaL < 107 )

Nu L = 0.15 Ra1/L 3 (10 7


< RaL < 1011 )

Como que h varia com L quando Nu L Ra1L 3


Face inferior aquecida ou face superior arrefecida

Ts > T Ts < T

Nu L = 0.27 Ra1/L 4 (10 5


< RaL < 1010 )

 Por que razo estas condies conduzem a uma menor taxa de


transmisso de calor do que as do slide anterior?
Problem 9.31: Convection and radiation losses from the surface of a
central solar receiver.

KNOWN: Dimensions and emissivity of cylindrical solar receiver. Incident solar flux.
Temperature of ambient air.

FIND: (a) Heat loss and collection efficiency for a prescribed receiver temperature, (b) Effect
of receiver temperature on heat losses and collector efficiency.

ASSUMPTIONS: (1) Steady-state, (2) Ambient air is quiescent, (3) Incident solar flux is
uniformly distributed over receiver surface, (4) All of the incident solar flux is absorbed by the
receiver, (5) Negligible irradiation from the surroundings, (6) Uniform receiver surface
temperature, (7) Curvature of cylinder has a negligible effect on boundary layer development,
(8) Constant properties
PROPERTIES: Table A-4, air (Tf = 550 K): k = 0.0439 W/mK, = 45.6 10-6 m2/s, =
66.7 10-6 m2/s, Pr = 0.683, = 1.82 10-3 K-1.

ANALYSIS: (a) The total heat loss is

q = q rad + q conv = As Ts4 + hAs ( Ts T )

With RaL = g (Ts - T)L3/ = 9.8 m/s2 (1.82 10-3 K-1) 500K (12m)3/(45.6 66.7 10-12
m4/s2) = 5.07 1012, the Churchill and Chu correlation yields
2

k
1/ 6
0.387 Ra L 0.0439 W / m K
h = 0.825 +
8 / 27
= {0.825 + 42.4} = 6.83 W / m 2 K
2
L 1 + ( 0.492 / Pr )9 / 16 12m

Hence, with As = DL = 264 m2

q = 264 m 2 0.2 5.67 108 W / m 2 K 4 ( 800 K ) + 264 m 2 6.83 W / m 2 K ( 500 K )


4

q = q rad + q conv = 1.23 106 W + 9.01 105 W = 2.13 106 W


With As q s = 2.64 107 W, the collector efficiency is

As qs q
= 100 =
(
2.64 107 2.13 106 W
(100 ) = 91.9%
)
A q
s s 7
2.64 10 W

(b) As shown below, because of its dependence on temperature to the fourth power, qrad
increases more significantly with increasing Ts than does qconv, and the effect on the efficiency is
pronounced

5E6 100

4E6
95

Collector efficiency, %
Heat rate, W

3E6
90
2E6

1E6 85

0 80
600 700 800 900 1000
Receiver temperature, K
75
Convection 600 700 800 900 1000
Radiation
Total Receiver temperature, K

COMMENTS: The collector efficiency is also reduced by the inability to have a perfectly
absorbing receiver. Partial reflection of the incident solar flux will reduce the efficiency by at
least several percent.
Cilindro horizontal
Desenvolvimento da camada limite e variao do nmero de Nusselt
local para um cilindro aquecido:

Nmero de Nusselt mdio:


2

0.387 Ra1/D 6
Nu D = 0.60 + RaD < 1012
9 /16 8 / 27
(
1 + 0.559 / Pr )

Como variam as condies para um cilindro arrefecido?


Esferas
Nmero de Nusselt mdio:

0.589 Ra1/D 4
Nu D = 2 + 4/9
1 + ( 0.469 / Pr )9 /16

 O que sucede quando RaD 0 ?


Conveco entre placas paralelas

L/S pequeno: camadas limites no chegam a coalescer e cada placa


comporta-se como se estivesse isolada

L/S elevado: h interaco entre camadas limites


Conveco entre placas paralelas
Correlaes de Elenbaas

a) Placas isotrmicas mesma temperatura, Ts


34
1 S 35
Nu s = Ra s 1 exp
24 L Ra s S L
q A S g (Ts T ) S 3
Nu s = Ras =
Ts T k
No limite de escoamento completamente desenvolvido, S/L  0:

1 S
Nu s , fd = Ra s
24 L

b) Uma placa isotrmica temperatura Ts,1 e a outra isolada; para a placa


isotrmica tem-se

1 S
Nu s , fd = Ra s
12 L
Conveco entre placas paralelas

c) Placas com fluxo constante e igual nas superfcies:


12
S
Nu s , L , fd = 0.144 Ra *s
L
qs S g qs S 4
Nu s , L = Ra =
*
Ts , L T k s
k

d) Uma placa com fluxo fluxo constante e a outra isolada:

12
S
Nu s , L , fd = 0.204 Ra *s
L
Conveco entre placas paralelas
Correlaes de Bar.Cohen e Rohsenow:
1 2
(a) Condies isotrmicas C1 C2 Ts + T
Nu s = + 12 T =
Casos (i) e (iii) (Ra s S L ) (Ra s S L )
2
2
1 2
(b) Condies isotrmicas C C2 Ts , L + T
Nu s = * 1 + 25 T =
Casos (ii) e (iv) Ra s S L (Ra s S L ) 2

Caso Condies de fronteira C1 C2 Sopt Smax/Sopt

576 2.87 2.71 (Ras S 3 L )


Placas simtricas isotrmicas, 1 4
(i) 1.71
Ts,1=Ts,2

2.51 2.12 (Ras* S 4 L )1 5


Placas com fluxo constante
(ii) qs,1 = qs, 2 48 4.77

(iii) Uma placa isotrmica e uma isolada 144 2.87 2.15 Ra s (


S 3
L )1 4
1.71

2.51 2.51 (Ras S 4 L )


Uma placa com fluxo constante e 1 5
(iv) uma isolada 24 4.77
Placas isotrmicas

S diminui Nu s diminui, mas n placas pode aumentar


Logo, existe Sopt que maximiza a taxa de transmisso de calor

Smax a distncia entre placas que maximiza o calor trocado em


cada placa

Placas com fluxo constante

S diminui diminui a taxa de t.c. por unidade de volume; Ts aumenta


Como Ts no pode aumentar indefinidamente, existe Sopt que maximiza
a taxa de t.c. por unidade de diferena de temperatura Ts(L) - T

Smax a distncia entre placas que, para um dado fluxo, minimiza a


temperatura da superfcie
Cavidades
Cavidades Rectangulares

 Paredes opostas a temperaturas diferentes e restantes paredes


perfeitamente isoladas
g (T1 T2 ) L3
 RaL

q = h (T1 T2 )
 Cavidade horizontal = 0, 180deg
 Cavidade vertical = 90 deg
Cavidades horizontais
 Aquecimento na base ( = 0 )
RaL < RaL , c = 1708 :
Camada de fluido termicamente estvel
Nu L = hL = 1
k
1708 < Ra L < 5 104 :

Instabilidade trmica provoca correntes de conveco regulares de


forma celular

3 105 < RaL < 7 109 :


O escoamento passa a turbulento
Nu L = 0.069 Ra1/L 3 Pr 0.074
 Aquecimento no topo ( = 180)
Camada de fluido incondicionalmente estvel
Nu L = 1

Cavidades verticais ( = 90)

 RaL < 10 :
3

Nu L = 1

 RaL > 10 :
3

Forma-se uma clula primria, com a


velocidade na regio central da cavidade
cada vez menor, e desenvolvem-se clulas
secundrias junto aos cantos medida que
RaL aumenta
 Correlations for Nu L Eqs. (9.50) - (9.53).

 Correlaes para Nu L  ver Eqs. (9.50) (9.53) do livro de Incropera e


de Witt
Cavidades inclinadas
 Relevante para colectores solares planos

 A taxa de transmisso de calor depende do ngulo de inclinao


relativamente a um ngulo de inclinao crtico *, cujo valor
funo de H/L (Tabela 9.4).

 A taxa de transmisso de calor depende tambm de RaL relativo a um


valor crtico RaL,c =1708/cos .

 Correlaes: Eqs. (9.54) (9.57).


Cavidades anulares
Cilindros concntricos

2 keff
 q = (Ti To )
1n ( Do Di )

 keff: condutibilidade trmica efectiva


 Numero de Rayleigh crtico:
1n ( Do / Di )
4

Rac* = RaL
L (D
3
i
3 / 5 3 / 5 5
+D
o )
L ( Do Di ) / 2
 Rac < 100 :
*

keff / k = 1

 100 < Rac < 10 :


* 7

( )
1/ 4
( Rac* )
keff Pr 1/ 4
= 0.386
k 0.861 + Pr
Esferas concntricas
DD
 q = keff i o (Ti To )
L
 Nmero de Rayleigh crtico:

Ra
Ras* = L L
( Do / Di ) ( D
4 7 / 5
+ Do )
7 / 5 5
i

 Ras < 100 : keff / k = 1


*

 100 < Ras < 10 :


* 4

( )
1/ 4
( Ras* )
keff Pr 1/ 4
= 0.74
k 0.861 + Pr
Regime misto conveco forada conveco natural

 Os efeitos de conveco forada e natural so ambos importantes se


(GrL )
Re 2L ~ O(1)

 O efeito de conveco natural dominante se (Gr


L )
Re 2L >> O(1)

 O efeito de conveco forada dominante se (Gr L )


Re 2L << O(1)

 Correlaes para transmisso de calor por conveco em regime misto

Nu n Nu FC
n
Nu NC
n

+ : Fora de impulso actua no mesmo sentido ou


perpendicularmente ao escoamento
- : Fora de impulso actua no sentido oposto ao do escoamento

n3
Problem 9.74: Use of saturated steam to heat a pharmaceutical in a batch reactor.

KNOWN: Volume, thermophysical properties, and initial and final temperatures of a


pharmaceutical. Diameter and length of submerged tubing. Pressure of saturated steam
flowing through the tubing.

FIND: (a) Initial rate of heat transfer to the pharmaceutical, (b) Time required to heat the
pharmaceutical to 70C and the amount of steam condensed during the process.
SCHEMATIC:

ASSUMPTIONS: (1) Pharmaceutical may be approximated as an infinite, quiescent fluid of


uniform, but time-varying temperature, (2) Free convection heat transfer from the coil may be
approximated as that from a heated, horizontal cylinder, (3) Negligible thermal resistance of
condensing steam and tube wall, (4) Negligible heat transfer from tank to surroundings, (5)
Constant properties.

PROPERTIES: Table A-4, Saturated water (2.455 bars): Tsat = 400K = 127C, hfg = 2.183
106 J/kg. Pharmaceutical: See schematic.

ANALYSIS: (a) The initial rate of heat transfer is q = hAs ( Ts Ti ) , where As = DL = 0.707
m2 and h is obtained from Eq. 9.34.
With = /Pr = 4.0 10-7 m2/s and RaD = g (Ts Ti) D3/ = 9.8 m/s2 (0.002 K-1) (102K)
(0.015m)3/16 10-13 m4/s2 = 4.22 106,
2 2
( )
6 1/ 6
0.387 Ra D1/ 6 0.387 4.22 10

Nu D = 0.60 + = 0.60 + = 27.7
8 / 27 8 / 27
1 + ( 0.559 / Pr )9 /16 1 + ( 0.559 /10 )9 /16


Hence, h = Nu D k / D = 27.7 0.250 W / m K / 0.015m = 462 W / m 2 K

and q = hAs ( Ts Ti ) = 462 W / m 2 K 0.707 m 2 (102C ) = 33,300 W

(b) Performing an energy balance at an instant of time for a control surface about the liquid,

d ( c T )
= q ( t ) = h ( t ) As ( Ts T ( t ) )
dt

where the Rayleigh number, and hence h , changes with time due to the change in the
temperature of the liquid.
Integrating the foregoing equation numerically, the following results are obtained for the
variation of T and h with t.
75 470

Convection coefficient, hbar (W/m ^2.K)


Temperature, (C) 65 450

55 430

45 410

35 390

25 370
0 100 200 300 400 500 600 700 800 900 0 100 200 300 400 500 600 700 800 900
Time, t(s) Time, t(s)

The time at which the liquid reaches 70C is

t f 855s <

The rate at which T increases decreases with increasing time due to the corresponding
reduction in (Ts T), and hence reductions in Ra D , h and q.

The Rayleigh number decreases from 4.22 106 to 2.16 106, while the heat rate decreases
from 33,300 to 14,000 W.

The convection coefficient decreases approximately as (Ts T)1/3, while q ~ (Ts T)4/3.
The latent energy released by the condensed steam corresponds to the increase in thermal
energy of the pharmaceutical. Hence, m c h fg = c ( Tf Ti ) ,

and

c ( Tf Ti ) 1100 kg / m3 0.2 m3 2000 J / kg K 45C


mc = = = 9.07 kg <
h fg 6
2.183 10 J / kg

COMMENTS: (1) Over such a large temperature range, the fluid properties are likely to vary
significantly, particularly and Pr. A more accurate solution could therefore be performed if
the temperature dependence of the properties were known. (2) Condensation of the steam is a
significant process expense, which is linked to the equipment (capital) and energy (operating)
costs associated with steam production.
Heat Exchangers:
Design Considerations
Heat Exchanger Types
Heat exchangers are ubiquitous to energy conversion and utilization. They involve
heat exchange between two fluids separated by a solid and encompass a wide
range of flow configurations.

Concentric-Tube Heat Exchangers

Parallel Flow Counterflow

 Simplest configuration.

 Superior performance associated with counter flow.


Cross-flow Heat Exchangers

Finned-Both Fluids Unfinned-One Fluid Mixed


Unmixed the Other Unmixed

 For cross-flow over the tubes, fluid motion, and hence mixing, in the transverse
direction (y) is prevented for the finned tubes, but occurs for the unfinned condition.

 Heat exchanger performance is influenced by mixing.


Shell-and-Tube Heat Exchangers

One Shell Pass and One Tube Pass

 Baffles are used to establish a cross-flow and to induce turbulent mixing of the
shell-side fluid, both of which enhance convection.
 The number of tube and shell passes may be varied, e.g.:

One Shell Pass, Two Shell Passes,


Two Tube Passes Four Tube Passes
Compact Heat Exchangers
 Widely used to achieve large heat rates per unit volume, particularly when
one or both fluids is a gas.
 Characterized by large heat transfer surface areas per unit volume, small
flow passages, and laminar flow.

(a) Fin-tube (flat tubes, continuous plate fins)


(b) Fin-tube (circular tubes, continuous plate fins)
(c) Fin-tube (circular tubes, circular fins)
(d) Plate-fin (single pass)
(e) Plate-fin (multipass)
Overall Heat Transfer Coefficient
An essential requirement for heat exchanger design or performance calculations.

Contributing factors include convection and conduction associated with the


two fluids and the intermediate solid, as well as the potential use of fins on both
sides and the effects of time-dependent surface fouling.

With subscripts c and h used to designate the hot and cold fluids, respectively,
the most general expression for the overall coefficient is:

1 = 1 = 1
UA (UA )c (UA )h

1 Rf , c Rf , h 1
= + + Rw + +
(o hA)c (o A)c (o A)h (o hA)h
 Rf Fouling factor for a unit surface area (m K/W)
2

Table 11.1
 Rw Wall conduction resistance (K/W)

 o Overall surface efficiency of fin array (Section 3.6.5)


Af
o,c or h = 1 (1 f )
A c or h

A = At total surface area (fins and exposed base)


A f surface area of fins only

Assuming an adiabatic tip, the fin efficiency is

tanh ( mL )
f , c or h =
mL c or h

mc or h = ( 2U p / k wt )c or h


U p , c or h = h partial overall coefficient
1 + hR
f c or h
A Methodology for Heat Exchanger
Design Calculations
- The Log Mean Temperature Difference (LMTD) Method -
A form of Newtons Law of Cooling may be applied to heat exchangers by
using a log-mean value of the temperature difference between the two fluids:
q = U A T1m
T1 T2
T1m =
1n ( T1 / T2 )

Evaluation of T1 and T2 depends on the heat exchanger type.

Counter-Flow Heat Exchanger:

T1 Th,1 Tc ,1
= Th,i Tc , o

T2 Th ,2 Tc ,2
= Th , o Tc ,i
Parallel-Flow Heat Exchanger:

T1 Th,1 Tc ,1
= Th ,i Tc ,i

T2 Th ,2 Tc ,2
= Th, o Tc , o

 Note that Tc,o can not exceed Th,o for a PF HX, but can do so for a CF HX.
 For equivalent values of UA and inlet temperatures,
T1m ,CF > T1m, PF

Shell-and-Tube and Cross-Flow Heat Exchangers:


T1m = F T1m ,CF
F Figures 11.10 - 11.13
Overall Energy Balance
Application to the hot (h) and cold (c) fluids:

Assume negligible heat transfer between the exchanger and its surroundings
and negligible potential and kinetic energy changes for each fluid.

q = m h ( ih ,i ih , o )

q = m c ( ic , o ic ,i )

i fluid enthalpy
Assuming no l/v phase change and constant specific heats,
q = m h c p , h (Th ,i Th , o ) = Ch (Th ,i Th, o )

q = m c c p , c (Tc , o Tc ,i ) = Cc (Tc , o Tc ,i )

Ch ,Cc Heat capacity rates


Special Operating Conditions

 Case (a): Ch>>Cc or h is a condensing vapor ( Ch ) .


Negligible or no change in Th (Th , o = Th ,i ) .
 Case (b): Cc>>Ch or c is an evaporating liquid ( Cc ) .

Negligible or no change in Tc (Tc , o = Tc ,i ) .


 Case (c): Ch=Cc.
T1 = T2 = T1m
Heat Exchangers:
The Effectiveness NTU Method
General Considerations
Computational Features/Limitations of the LMTD Method:

 The LMTD method may be applied to design problems for


which the fluid flow rates and inlet temperatures, as well as
a desired outlet temperature, are prescribed. For a specified
HX type, the required size (surface area), as well as the other
outlet temperature, are readily determined.
 If the LMTD method is used in performance calculations for which
both outlet temperatures must be determined from knowledge of the
inlet temperatures, the solution procedure is iterative.

 For both design and performance calculations, the effectiveness-NTU


method may be used without iteration.
Definitions

Definitions
Heat exchanger effectiveness, :

q
=
qmax

0 1

Maximum possible heat rate:

qmax = Cmin (Th,i Tc ,i )

Ch if Ch < Cc
Cmin = or
Cc if Cc < Ch

 Will the fluid characterized by Cmin or Cmax experience the largest possible
temperature change in transit through the HX?

 Why is Cmin and not Cmax used in the definition of qmax?


Number of Transfer Units, NTU
NTU UA
Cmin
 A dimensionless parameter whose magnitude influences HX performance:
q with NTU
Heat Exchanger Relations

q =

m (
h ih , i ih , o )
or
q = Ch (Th,i Th , o )


q = mc ( ic , o ic ,i )


or
q = Cc (Tc , o Tc ,i )

q = Cmin (Th ,i Tc ,i )

Performance Calculations:
 = f ( NTU , Cmin / Cmax )

Cr
 Relations Table 11.3 or Figs. 11.14 - 11.19
-NTU Expressions (Table 2.2 of the book, more detail in book)
Type of HEX (NTU,C*) NTU(,C*)

[( ) ]

1 exp 1 C NTU 1 1 C
ln
1
[( ) ]
= NTU =
Counterflow
1 C exp 1 C NTU 1C

Parallel Flow
=
1

[1 exp[ (1 + C )NTU ]]

NTU =
1

[ (
ln 1 + 1 + C

)]
1+ C 1+C

1 exp( C NTU )
[ )]
Cross flow, Cmin
mixed and Cmax = 1 exp
NTU =
1


ln 1 + C ln 1 (
unmixed C C

Cross flow, Cmax


mixed and Cmin
1
[
{ [
= 1 exp C 1 exp NTU ( )]}] NTU = -ln 1 +


1

(
ln 1 C

)
unmixed C
C

=
2 1 / 2
2 1 + C 1 + C 2
1 / 2
1 to 2 shell-and- 1 + exp NTU 1 + C 2 NTU =
1
( )

(1 + C 2 )1/ 2
ln
2 1 / 2 1 / 2
tube HEX 1+C + 1+ C
1 / 2 2 1 + C + 1 + C 2
1 exp NTU 1 + C 2


Design Calculations:
NTU = f ( , Cmin / Cmax )

 Relations Table 11.4 or Figs. 11.14 - 11.19

For all heat exchangers,


with Cr

For Cr = 0, a single NTU relation applies to all HX types.

= 1 exp ( NTU )
or

NTU = 1n (1 )
Radiao: Consideraes gerais
Estuda-se radiao trmica, cujas origens esto ligadas emisso da matria a
uma temperatura absoluta T>0
A emisso devida oscilaes e transies electrnicas dos muitos electres que
constituem a matria que, por sua vez, so mantidos pela energia trmica da matria
A emisso corresponde energia transferida da matria (calor) e, portanto, corresponde
reduo de energia trmica armazenada na matria.
A radiao tambm pode ser intersectada e absorvida pela matria.
A absoro resulta em transferncia de calor para a matria e, portanto, corresponde
a um amento de energia trmica armazenada na matria.
Considere um slido temperatura Ts num recinto
fechado com vcuo, cujas paredes esto temperatura Tsur

 Que fenmeno ocorre se Ts > Tsur? Porqu


 Que fenmeno ocorre se Ts < Tsur? Porqu
Radiao: Consideraes gerais
A emisso de gases ou slidos semi-transparentes ou lquidos um
fenmeno volumtrico.
A emisso de slidos ou lquidos opacos um fenmeno superficial (com a
emisso originria em tomos ou molculas a 1 m da superfcie).

A natureza dual da radiao:


Nalguns casos, as manifestaes fsicas da radiao podem ser explicadas
olhando-as como partculas (aka fotes ou quanta).
Noutros casos, a radiao comporta-se como uma onda electromagntica.
Radiao: Consideraes gerais

Em qualquer dos casos, a radiao caracterizada por um comprimento de onda


e frequncia que esto relacionados pela velocidade de propagao da radiao
no meio em causa, c: = c

No vcuo: c = co = 2.998 x 108 m/s

O espectro electromagntico
Radiao: Consideraes gerais

A quantidade de radiao emitida por uma superfcie opaca varia com o


comprimento de onda, podendo falar-se em distribuio espectral em todos os
comprimentos de onda ou de componentes monocromticas/espectrais associadas a
comprimentos de onda especficos.
Radiao: Consideraes gerais

Efeitos direccionais

A radiao emitida por uma superfcie s-lo- em todas as


direces do hemisfrio e segundo uma distribuio direccional

A direco pode ser representada em coordenadas esfricas pelo


ngulo polar ou zenital e pelo ngulo azimutal .
Radiao: Consideraes gerais

Intensidade de radiao

A quantidade de radiao emitida por uma superfcie, dA1, e que se


propaga numa direco particular, (,), quantificada em termos
de um ngulo slido diferencial associado direco em causa.

dAn
d
r2

dAn elemento unitrio de superfcie de uma esfera hipottica na direco (,),


Radiao: Consideraes gerais
Intensidade de radiao

dAn = r 2 sin d d

dAn
d = 2
= sin d d
r

O ngulo slido tem unidades steradianos (sr).


2 /2
hemi = 0 0 sin d d = 2 sr

Intensidade Espectral : a quantidade usada para especificar o fluxo de energia radiante


(W/m2) num ngulo slido unitrio numa direco prescrita (W/m2.sr) e num intervalo
unitrio de comprimentos de onda (W/m2.sr.m).
Radiao: Consideraes gerais
Intensidade de radiao
A intensidade espectral, I ,e , associada emisso de um elemento de rea unitrio, dA1,
num ngulo slido, d (em torno de e ),e num intervalo de comprimento de onda, d ,
(em torno de ), :
dq
I ,e ( , , )
( dA1 cos ) d d
O argumento para definir o fluxo radiativo em termos da rea projectada da superfcie
( dA1 cos ) emerge do facto de haver superfcies para as quais, com boa aproximao,
I ,e independente da direco: superfcies difusas, e a radiao isotrpica.

A rea projectada como dA1 apareceria


Se observada segundo os ngulos ,

Quanto vale a rea projectada para = 0 ?

Quanto vale a rea projectada para = / 2 ?


Radiao: Consideraes gerais
Intensidade de radiao
A taxa de calor espectral e o fluxo de calor espectral associados emisso a partir de
dA1 so, respectivamente,

dq
dq = I ,e ( , , ) dA1 cos d
d

dq = I ,e ( , , ) cos d = I ,e ( , , ) cos sin d d


Relao da intensidade com poder
emissivo, irradiao e radiosidade
O poder emissivo espectral (W/m2.m) corresponde emisso espectral em todas as
direces possveis: 2 / 2
E ( ) = 0 0 I ,e ( , , ) cos sin d d

O poder emissivo total (W/m2) corresponde emisso espectral em todas as


direces e comprimentos de onda possveis:
E= 0 E ( ) d

Para superfcies difusas, a emisso isotrpica e: E ( ) = I ,e ( ) E = I e

A intensidade espectral da radiao incidente numa superfcie, I ,i


definida em termos do ngulo slido unitrio em torno da
direco de incidncia, do intervalo de comprimento de onda, d
em torno de, , e da rea projectada do receptor, dA1 cos .
Relao da intensidade com poder
emissivo, irradiao e radiosidade
A irradiao espectral ( W/m 2
m ) vale:
2 /2
G ( ) = 0 0 I ,i ( , , ) cos sin d d

e a irradiao total ( W/m 2 ) :



G= 0 G ( ) d

 Quantos G e G so expressos se a radiao for difusa?

A radiosidade de uma superfcie opaca contabiliza


toda a radiao que abandona a superfcie em todas
as direces e pode incluir as contribuies da
reflexo e emisso.
Relao da intensidade com poder
emissivo, irradiao e radiosidade
Com I ,e + r a designar a intensidade espectral associada radiao emitida pela superfcie
e a reflexo da radiao incidente, a radiosidade espectral ( W/m 2 m ) :

2 /2
J ( ) = 0 0 I ,e + r ( , , ) cos sin d d

E a radiosidade total ( W/m 2 ) :


J = 0 J ( ) d

 Quantos J e J podem ser expressos se a superfcie emitir e reflectir de forma difusa?


Radiao de CORPO NEGRO
O Corpo Negro
 Uma idealizao que fornece os limites da radiao emitida e absorvida
pela matria.

Para uma dada temperatura e um dado comprimento de onda,


nenhuma superfcie capaz de emitir mais que um corpo negro: emissor
ideal.

Um corpo negro um emissor difuso.

Um corpo negro absorve toda a radiao incidente: absorvedor ideal.


Radiao de CORPO NEGRO
A Cavidade Isotrmica (Hohlraum):
parede interior a temperatura uniforme.

(a) Depois de mltiplas reflexes, toda a


radiao entrada na cavidade
virtualmente absorvida corpo negro.

(b) A emisso a partir da abertura a mxima que se pode atingir, para a


temperatura associada cavidade, e difusa (independente da direco)
corpo negro.
(c) O campo radiativo no interior da cavidade (efeito acumulado da radiao emitida
e da reflectida pela parede da cavidade) tem por efeito assegurar uma irradiao
difusa (correspondente emisso de um corpo negro numa forma igual radiao emergente
pela abertura). O campo radiativo no interior da cavidade de corpo negro.
Qualquer superfcie no interior da cavidade irradiada de forma difusa: ( G = E ,b )

Esta condio depende da superfcie da cavidade ser acentuadamente


reflectora ou absorvedora?
Lei da distribuio de Plank
A distribuio espectral do poder emissivo de um corpo negro (determinado teoricamente
e confirmado experimentalmente) (Plank):
C1
E ,b ( ,T ) = I ,b ( ,T ) =
5 exp ( C2 / T ) 1

Primeira constante:
C1 = 3.742 x 108 W m 4 / m 2
Segunda constante: C2 = 1.439 x 104 m K

 E ,b varia continuamente com e aumenta com T.

 A distribuio caracterizada por um mximo para o qual max dado


pela lei deslocamento de Wien :
maxT = C3 = 2898 m K
 A quantidade fraccional da emisso total de corpo negro que aparece a baixos
comprimentos de onda aumenta com o aumento de T.
Lei de Stefan-Boltzmann
O poder emissivo total de um corpo negro obtido integrando a distribuio de Planck
em todos os comprimentos de onda possveis.
Eb = I b = 0 E ,b d = T 4

a lei de Stefan-Boltzmann, em que:

a constante se Stefan-Boltzmann, = 5,670 10-8 W/m2.K4

A fraco total da emisso de um corpo negro que est contida num intervalo de comprimento

de onda prescrito ou banda ( 1 < < 2 )K :


02 E ,b d o1 E ,b d
F( 1 2 ) = F( 0 2 ) F( 0 1 ) =
T 4


E d
F( 0 ) = 0 ,b = f ( T )
T
Lei de Stefan-Boltzmann fraco de
energia

C1
E ,b ( , T ) = I ,b ( ,T ) =
exp ( C2 / T ) 1
5


K
Band Emission (cont)

Note ability to readily determine I ,b and its relation to the maximum intensity from
the 3rd and 4th columns, respectively.
 If emission from the sun may be approximated as that from a blackbody at
5800K, at what wavelength does peak emission occur?
 Would you expect radiation emitted by a blackbody at 800K to be discernible
by the naked eye?
 As the temperature of a blackbody is increased, what color would be
the first to be discerned by the naked eye?
Problem 12.6: Evaluation of total solar irradiation at the earths surface
from knowledge of the direct and diffuse components of
the incident radiation.

KNOWN: Flux and intensity of direct and diffuse components, respectively, of solar
irradiation.

FIND: Total irradiation.


SCHEMATIC:

ANALYSIS: Since the irradiation is based on the actual surface area, the contribution due to
the direct solar radiation is
cos .
G dir = q dir

For the contribution due to the diffuse radiation


G dif = Idif .

Hence
cos + Idif
G = G dir + G dif = q dir
or

G = 1000 W / m 2 0.866 + sr 70 W / m 2 sr

G = ( 866 + 220 ) W / m 2

G = 1086 W / m 2 .

COMMENTS: Although a diffuse approximation is often made for the non-direct component
of solar radiation, the actual directional distribution deviates from this condition, providing
larger intensities at angles close to the direct beam.
Problem 12.18: Determination of the suns emissive power, temperature
and wavelength of maximum emission, as well as the
earths temperature, from knowledge of the sun/earth
geometry and the solar flux at the outer edge of the earths
atmosphere.

KNOWN: Solar flux at outer edge of earths atmosphere, 1353 W/m2.

FIND: (a) Emissive power of sun, (b) Surface temperature of sun, (c) Wavelength of
maximum solar emission, (d) Earth equilibrium temperature.

SCHEMATIC:
ASSUMPTIONS: (1) Sun and earth emit as blackbodies, (2) No attenuation of solar
radiation enroute to earth, (3) Earth atmosphere has no effect on earth energy balance.

ANALYSIS: (a) Applying conservation of energy to the solar energy crossing two concentric
spheres, one having the radius of the sun and the other having the radial distance from the edge
of the earths atmosphere to the center of the sun, it follows that

( )
2
2 De
Es Ds = 4 R s e q s .
2
Hence

( )
2
4 1.5 10 m 0.65 10 m 1353 W / m 2
11 7
Es = = 6.302 107 W / m 2 .
(1.39 10 m )
9 2

(b) From the Stefan-Boltzmann law, the temperature of the sun is


1/ 4
E
1/ 4 6.302 107 W / m 2
Ts = s = = 5774 K.
8 2 4
5.67 10 W / m K

(c) From Wiens law, the wavelength of maximum emission is


C 2897.6 m K
max = 3 = = 0.50 m.
T 5774 K
(d) From an energy balance on the earths surface

( ) (
E e De2 = qS De2 / 4 . )
Hence, from the Stefan-Boltzmann law,
1/ 4 1/ 4
q 1353 W / m 2
Te = S = = 278 K.
4 5.67 108 W / m 2 K 4
4

COMMENTS: The average earth temperature is higher than 278 K due to the shielding effect
of the earths atmosphere (transparent to solar radiation but not to longer wavelength earth
emission).
Processos e propriedades radiativas.
Propriedades radiativas de superfcies
Emissividade de uma Superfcie
Introduzindo uma propriedade especfica, (a emissividade), a radiao emitida por uma
superfcie pode ser determinada, o que contrasta com o seu comportamento ideal como corpo
negro mesma temperatura.
A definio de emissividade depende de estarmos interessados em calcular o carcter
direccional e/ou espectral da radiao emitida, em contraste com mdias em todas as direces
(radiao hemisfrica) e/ou em todos os comprimentos de onda (total).
A emissividade espectral directional :
I ,e ( , , ,T )
, ( , , ,T )
I ,b ( ,T )

A emissividade espectral hemisfrica (uma mdia direccional):

E ( , T ) 02 0 / 2 I ,e ( , , ,T ) cos sin d d
( ,T ) =
E ,b ( , T ) 02 0 / 2 I ,b ( ,T ) cos sin d d
Processos e propriedades radiativas.
Propriedades radiativas de superfcies
A emissividade total hemisfrica (uma mdia direccional e espectral):

E (T ) 0 ( ,T ) E ,b ( ,T ) d ,
(T ) =
Eb (T ) Eb (T )
Com um boa de aproximao, a emissividade hemisfrica igual emissividade normal: = n

Valores tpicos da emissividade total normal :

Notas:
 Baixas emissividades dos metais polidos e
crescente emissividade de metais no polidos
e superfcies oxidadas.
 Emissividades comparativamente elevadas
dos no condutores.
Processos e propriedades radiativas.
Propriedades radiativas de superfcies
Variaes espectrais tpicas

Notar decrscimo de ,n com o


aumento de para metais
e comportamento diferente dos no metais

Variaes de temperatura tpicas :

Porque que n aumenta com o aumento de


para o tungstnio e no aumenta para o xido de alumnio?
Absoro, reflexo e transmisso:
resposta radiao incidente
Pode haver 3 respostas de um meio semi transparente irradiao:

 Reflexo pelo meio ( G ,ref ) .


 Absoro pelo meio ( G ,abs ) .
 Transmisso atravs do meio ( G ,tr ) .

Balano de Radiao

G = G ,ref + G ,abs + G ,tr

Contrastando com o que se disse para meios semitransparentes, efeitos volumtricos,


a resposta de um material opaco irradiao governado por fenmenos superficiais
e G ,tr = 0.
G = G ,ref + G ,abs + G ,tr

O comprimento de onda da radiao incidente e a natureza do material determinam se o


material semitransparente ou opaco.
 O vidro e a gua so semitransparentes ou opacos?
Absoro, reflexo e transmisso:
resposta radiao incidente
A menos que um material opaco esteja a uma temperatura suficientemente alta para emitir
radiao visvel, a sua cor determinada pela dependncia espectral da reflexo em resposta
irradiao visvel.

 O que se pode dizer sobre a reflexo de uma superfcie branca?


 E sobre uma negra?

 Porque que as folhas so verdes?


Absorsividade de uma superfcie opaca
A absorsividade espectral direccional, desprezando dependncia de T:
I ,i ,abs ( , , )
, ( , , )
I ,i ( , , )

A absorsividade espectral hemisfrica :


G ,abs ( ) 02 0 / 2 , ( , , ) I ,i ( , , ) cos sin d d
( ) =
G ( ) 02 0 / 2 I ,i ( , , ) cos sin d d
 Se a radiao for difusa, em que que se simplifica o resultado anterior?
E se a superfcie for difusa?

A absorsividade total hemisfrica : Gabs o ( ) G ( ) d
=
0 G ( ) d

G
 Se a irradiao for de corpo negro, como se escrevem as equaes anteriores?
 A absorsividade aproximadamente independente da temperatura da superfcie,
mas se a irradiao for de corpo negro, porque que depende da temperatura
do corpo negro?
Reflectividade de uma superfcie opaca
A reflectividade espectral direccional, desprezando dependncia de T:
I ,i ,ref ( , , )
, ( , , )
I ,i ( , , )

A absorsividade espectral hemisfrica :


/2
G ,ref ( ) 02 0 , ( , , ) I ,i ( , , ) cos sin d d
=
G ( ) I ,i ( , , )
 Se a radiao for difusa, em que que se simplifica o resultado anterior?
E se a superfcie for difusa?
A reflectividade total hemisfrica :

G ( ) G ( ) d
abs = 0
G 0 G ( ) d

Condies limitativas de reflexo difusa e


espectral. Superfcies polidas e rugosas.
Reflectividade de uma superfcie opaca

 Notar forte dependncia de (e = 1- ) em


 A neve uma substncia muito reflectora? E a tinta branca?
Transmissividade
A transmissividade espectral hemisfrica ,desprezando dependncia de T:

G ,tr

G ( )

Notar que a pequenos e elevados comprimentos de onda h mudana de


condies de semitransparente para opaco
Gtr 0 G ,tr ( ) d
=
0 G ( ) d
A reflectividade total hemisfrica :
G
Para um meio semitransparente,
+ + = 1 Para um meio opaco,
+ = 1
+ + = 1 + =1
Lei de Kirchhoff
A Lei de Kirchhoff estabelece que a emissividade total hemisfrica de uma
superfcie igual sua absorsividade total hemisfrica :
=
Contudo, as condies associadas sua derivao so muito restritivas:
A irradiao da superfcie corresponde emisso de um corpo negro mesma
temperatura do corpo.

Ainda assim, a lei de Kirchhoff pode aplicar-se s propriedades espectrais direccionais


sem restries:
, = ,

Porque que no h restries ao uso da equao anterior?


Supefcies difusas/cinzentas
Com 02 0 / 2 , cos sin d d
= 2 / 2
0 0 cos sin d d
e 02 0 / 2 , I ,i cos sin d d
= 2 / 2
0 0 I ,i cos sin d d
Em que condies se pode igualar a ?

Com 0 E ,b ( ) d
=
Eb (T )

e G ( ) d
= 0
G
Em que condies se pode igualar a ?

Condies associadas com a


hiptese de superfcie cinzenta
Problem 12.49: Determination of the solar absorptivity and total emissivity
of a diffuse surface from knowledge of the spectral
distribution of ( ) and the surface temperature.

KNOWN: Spectral, hemispherical absorptivity of an opaque surface.

FIND: (a) Solar absorptivity, (b) Total, hemispherical emissivity for Ts = 340K.
SCHEMATIC:

ASSUMPTIONS: (1) Surface is opaque, (2) = , (3) Solar spectrum has G = G,S
proportional to E,b (, 5800K).

ANALYSIS: (a) The solar absorptivity may be expressed as



S = ( ) E ,b ( , 5800K ) d / E ,b ( , 5800K ) d.
0 0

The integral can be written in three parts using F(0 ) terms.


S = 1 F( 0 0.3) + 2 F( 0 1.5) F( 0 0.3) + 3 1 F( 0 1.5) .

From Table 12.1,


T = 0.3 5800 = 1740 mK F(0 0.3 m) = 0.0335
T = 1.5 5800 = 8700 mK F(0 1.5 m) = 0.8805.
Hence,
S = 0 0.0355 + 0.9 [ 0.8805 0.0335] + 0.1[1 0.8805] = 0.774.

(b) The total, hemispherical emissivity for the surface at 340K may be expressed as

= ( ) E ,b ( ,340K ) d / E b ( 340K ) .
0

With = , the integral can be written in terms of the F(0 ) function. However, it is readily
recognized that since
F( 0 1.5 m, 340K ) 0.000 at T = 1.5 340 = 510 m K

there is negligible emission below 1.5 m.

It follows that
= = = 0.1

COMMENTS: The assumption = is satisfied if the surface is irradiated diffusely or if


the surface itself is diffuse. Note that for this surface under the specified conditions of solar
irradiation and surface temperature, S . Such a surface is spectrally selective.
Problem 12.90: Determination of the emissivity and absorptivity of a coated
vertical plate exposed to solar-simulation lamps and the magnitude
of the irradiation required to maintain a prescribed plate
temperature.

KNOWN: Vertical plate of height L = 2 m suspended in quiescent air. Exposed surface with
diffuse coating of prescribed spectral absorptivity distribution subjected to simulated solar
irradiation, GS,. Plate steady-state temperature Ts = 400 K.
FIND: Plate emissivity, , plate absorptivity, , and plate irradiation, G.

SCHEMATIC:

ASSUMPTIONS: (1) Steady-state conditions, (2) Ambient air is extensive, quiescent, (3)
Spectral distribution of the simulated solar irradiation, GS, , proportional to that of a blackbody at
5800 K, (4) Coating is opaque, diffuse, and (5) Plate is perfectly insulated on the edges and the
back side, and (6) Plate is isothermal.

PROPERTIES: Table A.4, Air (Tf = 350 K, 1 atm): = 20.92 10-6 m2/s, k = 0.030 W/mK,
= 29.90 10-6 m2/s, Pr = 0.700.
ANALYSIS: (a) Perform an energy balance on the plate as shown in the schematic on a per unit
plate width basis,

E& in E out = 0

G T 4 h ( T T ) L = 0
s s

where and are determined from knowledge of and h is estimated from an appropriate
correlation.

Plate total emissivity: Expressing the emissivity in terms of the band emission factor, F(0 - T),

= 1F( 0 T ) + 2 1 F( 0 T )
1 s 1 s

= 0.9 0 + 0.1[1 0 ] = 0.1 <

where, from Table 12.1, with ,Ts = 1m 400 K = 400 mK, F(0-T) = 0.000.
Plate absorptivity: With the spectral distribution of simulated solar irradiation proportional to
emission from a blackbody at 5800 K,

= 1F( 0 T ) + 2 1 F( 0 T )
1 s 1 s

= 0.9 0.7202 + 0.1[1 0.7202] = 0.676 <

where, from Table 12.1, with 1Ts = 5800 mK, F(0 -T) = 0.7202.

Estimating the free convection coefficient, h : Using the Churchill-Chu correlation with
properties evaluated at Tf = (Ts + T )/2 = 350 K,

g ( Ts T ) L3
Ra L =

9.8 m s 2 (1 350 K ) 100 K ( 2 m )


3
Ra L = = 3.581 1010
6 2 6 2
20.92 10 m s 29.90 10 m s
2

0.387Ra1/
L
6
Nu L = 0.825 + =377.6
8 27
1 + ( 0.492 Pr )9 16

h L = Nu L k L = 377.6 0.030 W m K 2 m = 5.66 W m 2 K <


Irradiation on the Plate: Substituting numerical values into Eq. (1),
4
0.676G 0.1 5.67 108 W m 2 K ( 400 K )4 5.66 W m 2 K ( 400 300 ) K = 0

G = 1052 W m 2
Solar Radiation
The sun is a nearly spherical source of radiation whose outer diameter is
1.39 x 109 m and whose emissive power approximates that of a blackbody at 5800K.

The distance from the center of the sun to the center of the earth varies with time
of year from a minimum of 1.471 x 1011 m to a maximum of 1.521 x 1011 m, with
an annual average of 1.496 x 1011 m.

Due to the large sun-to-earth distance, the suns rays


are nearly parallel at the outer edge of the earths
atmosphere, and the corresponding radiation flux is
qS = f x Sc

Sc the solar constant or heat flux (1353 W/m 2 )


when the earth is at its mean distance from the sun.

f correction factor accounting for eccentricity


of the earth's orbit ( 0.97 < f <1.03)
Extraterrestrial irradiation of a surface whose normal is at a zenith angle
relative to the suns rays is
GS ,o = f x Sc x cos

Interaction of solar radiation with earths atmosphere:


 Absorption by aerosols over the entire spectrum.
 Absorption by gases (CO2, H2O ( v ), O3) in discrete wavelength bands.
 Scattering by gas molecules and aerosols.
Effect of Atmosphere on Spectral Distribution of Solar Radiation:

 Attenuation over the entire spectrum but more pronounced in spectral bands
associated with polar molecules.
 Note concentration of all radiation in the spectral region 0.3 < < 3 m and
peak at 0.5 m.
 Why is the assumption of graybody behavior often inappropriate for
surfaces experiencing solar irradiation?
Effect of Atmosphere on Directional Distribution of Solar Radiation:

 Rayleigh scattering is approximately uniform in all directions (isotropic


scattering), while Mie scattering is primarily in the direction of the suns
rays (forward peaked).

 Directional distribution of radiation


at the earths surface has two components.
Direct radiation: Unscattered and in
the direction of the suns rays.
Diffuse radiation: Scattered radiation
strongly peaked in the forward direction.
 Calculation of solar irradiation for a
horizontal surface often presumes the
scattered component to be isotropic.
cos + I dir
GS = GS ,dir + GS ,dif = qdir

0.1 < ( GS ,dir / GS ) < 1.0


Clear skies Completely overcast
Terrestrial Radiation
Emission by Earths Surface:
E = T 4

 Emissivities are typically large. For example, from Table A.11:


Sand/Soil: > 0.90
Water/Ice: > 0.95
Vegetation: > 0.92
Snow: > 0.82
Concrete/Asphalt: > 0.85

 Emission is typically from surfaces with temperatures in the range of


250 < T < 320K and hence concentrated in the spectral region
4 < < 40 m, with peak emission at 10 m.

Atmospheric Emission:
 Largely due to emission from CO2 and H2O (v) and concentrated in the
spectral regions 5 < < 8 m and > 13 m.
 Although far from exhibiting the spectral characteristics of blackbody emission,
earth irradiation due to atmospheric emission is often approximated by a
blackbody emissive power of the form
Gatm = Tsky
4

Tsky the effective sky temperature

230K< Tsky < 285K


Cold, clear sky Warm, overcast sky

Can water in the natural environment freeze if the ambient air temperature
exceeds 273K? If so, what environmental conditions (wind and sky)
favor ice formation?
Surface Radiative Properties
Concentration of solar ( 0.3 < < 3 m ) and terrestrial ( 4 < < 40 m ) in
different spectral regions often precludes use of the gray surface approximation
( S ) .

 Note significant differences in and for the two spectral regions: snow,
human skin, white paint.
 In terms of net radiation transfer to a surface with solar irradiation, the parameter
S / has special significance. Why?
Surface S /
Snow 0.29

Human skin 0.64 Rejection

White paint 0.22

Black paint 1.0


Collection
Evaporated Al film 3.0
Problem 12.119: Determination of preferred roof coating (Parsons Black,
Acrylic White, or Zinc Oxide White) and corresponding
heat load for prescribed operating conditions.

KNOWN: Dimensions and construction of truck roof. Roof interior surface temperature. Truck
speed, ambient air temperature, and solar irradiation.

FIND: (a) Preferred roof coating, (b) Roof surface temperature, (c) Heat load through roof,
(d) Effect of velocity on surface temperature and heat load.
SCHEMATIC:

ASSUMPTIONS: (1) Turbulent boundary layer development over entire roof, (2) Constant
properties, (3) Negligible atmospheric (sky) irradiation, (4) Negligible contact resistance.

PROPERTIES: Table A.4, Air (Ts,o 300 K, 1 atm): = 15 106 m 2 s , k = 0.026 W m K ,


Pr = 0.71.

ANALYSIS: (a) To minimize heat transfer through the roof, minimize solar absorption relative
to surface emission. Hence, from Table A.12, use zinc oxide white for which S = 0.16
and = 0.93.


(b) Performing an energy balance on the outer surface of the roof, S GS + q conv
E q cond = 0,
it follows that
4
S GS + h(T Ts,o ) = Ts,o + (k t)(Ts,o Ts,i )
where it is assumed that convection is from the air to the roof. With
VL 30 m s(5 m)
Re L = = = 107
6 2
15 10 m s

Nu L = 0.037 Re 4L / 5 Pr1/ 3 = 0.037(107 ) 4 / 5 (0.71)1 / 3 = 13,141

h = Nu L (k L) = 13,141(0.026 W m K/5 m = 68.3 W m 2 K .

Substituting numerical values in the energy balance and solving by trial-and-error, we obtain

Ts,o = 295.2 K.

(c) The heat load through the roof is

q = (kAs t)(Ts,o Ts,i ) = (0.05 W m K 10 m 2 0.025 m)35.2 K = 704 W .

(d) From parametric calculations based on the foregoing model, the following results are
obtained.
300 700

295 650
Temperature, Tso(K)

Heat load, q(W)


290 600

285 550

500
280
5 10 15 20 25 30
5 10 15 20 25 30
Velocity, V(m/s)
Velocity, V(m/s)

The surface temperature and heat load decrease with decreasing V due to a reduction in the
convection heat transfer coefficient and hence convection heat transfer from the air.

COMMENTS: The heat load would increase with increasing S/.


Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
Conceitos bsicos
Recinto fechado consiste de 2 ou mais superfcies que englobam uma regio do espao
(tipicamente preenchida com gs) e que trocam energia radiativa entre si.

Um meio no participativo, num recinto fechado, no emite, no absorve,


nem sofre scattering de energia radiativa. Portanto, no produz qualquer efeito nas
trocas de radiao entre as superfcies.

Cada superfcie que limita o recinto fechado suposta ser isotrmica, opaca, difusa
e cinzenta, sendo caracterizada por radiosidade e irradiao uniformes.
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
O Factor de Forma
O factor de forma, Fij , , uma quantidade geomtrica correspondente fraco da
radiao que abandona a superfcie i e que intersectada pela superfcie j.

qi j
Fij =
Ai J i

Considere a troca entre duas reas diferenciais:

cosi cos j
dqi j = I i cosi dAi d j i = J i dAi dAj
R2

cosi cos j
Fij = 1 A Aj dAi dAj
Ai i R 2
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
Relaes para o Factor de Forma
cosi cos j
Relao de Reciprocidade:
F ji = 1 Ai A j dAi dA j
Aj R 2

Ai Fij = A j F ji

Regra da Somatrio para recintos fechados.

N
Fij = 1
j =1
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
Relaes para o Factor de Forma
Geometrias Bi-Dimensionais (Tabela 13.1). Por exemplo,

Um Plano Infinito e uma


Fileira de Cilindros

( ) ( )
2 1/ 2
2 1/ 2
+ D tan1 s 2D
2
Fij = 1 1 D
s s D
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
Relaes para o Factor de Forma
Geometrias Tri-Dimensionais (Tabela 13.2). Por exemplo,

Discos Coaxiais Paralelos


Fij = 1 S S 2 4 ( rj / ri )
2 1/ 2


2
1 + R 2j
S = 1+
Ri2

Ri = ri / L R j = rj / L
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
FACTORES DE FORMA MTODO DAS CORDAS
PERMITE CALCULAR REAS DE PERMUTA, AiFij, ENTRE 2 SUPERFCIES i E j QUE
OBEDECEM S SEGUINTES CONDIES:

A1

A2

COMPRIMENTO MUITO MAIOR QUE A DISTNCIA QUE AS SEPARA


SECES RECTAS CONSTANTES E PERPENDICULARES AO COMPRIMENTO
DISTNCIA QUE AS SEPARA SER CONSTANTE
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
FACTORES DE FORMA MTODO DAS CORDAS

L3
L1

L2

L4

A1F12 =
cordas cruzadas cordas no cruzadas
2

A1F12 =
(L1 + L2 ) (L1 + L2 )
2
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
FACTORES DE FORMA ENTRE RECTNGULOS DIAMETRALMENTE OPOSTOS
cosi cos j
dq =
APLICA-SE QUANDO H SIMETRIA OU SEMELHANA
i j I cosNAdAd =
i GEOMETRIA
j i J dAdA
i i i
R 2 i j

A1 cosi cos j
Fij = 1 A Aj dAdA
Ai i R 2 i j

R
A3

A1F12 = A3 F34
A4

A2
Trocas radiativas entre superfcies: recintos
fechados com meio no participativo
FACTORES DE FORMA OUTRAS RELAES

n
Fi ( j ) = Fik
k =1

n
A( j )F( j )i = Ak Fki
k =1
n

A F k ki
F( j )i = k =1

A( j )
Trocas radiativas entre superfcies negras
Para um corpo negro, J i = Ebi .

Troca de calor til entre duas superfcies


que podem ser aproximadas como corpos
negros taxa til qual a radiao
abandona a superfcie i devido
sua interaco com j

ou taxa til qual a superfcie j


qij = qi j q j i
ganha radiao devido qij = Ai Fij Ebi A j F ji Ebj
qij = Ai Fij (Ti 4 T j4 )
sua interaco com i

Transferncia de radiao til da superfcie i devido a trocas com todas as (N)


superfcies num recinto fechado:
qi = Ai Fij (Ti 4 T j4 )
N

j =1
Troca radiativa entre as N superfcies opacas,
difusas e cinzentas de um recinto fechado

Expresses alternativas para transferncia radiativa


til a partir da superfcie i:

qi = Ai ( J i Gi ) Fig. (b) (1)

qi = Ai ( Ei iGi ) Fig. (c) (2)

Ebi J i
qi = Fig. (d) (3)
( i) i i
1 / A

Sugere uma resistencia da superfcie


radiativa da forma: (1 i ) / i Ai
Troca radiativa entre as N superfcies opacas,
difusas e cinzentas de um recinto fechado
Ji J j
qi = Ai Fij ( J i J j ) =
N N

(AF )
1 (4)
j =1 j =1
i ij

Sugere uma resistncia espacial


(
ou geomtrica da forma: A F 1
i ij )
Igualando as Eqs. (3) e (4) corresponde a um balano de energia radiativa
superfcie i:
Ebi J i N J J
= i j
(5)
(1 i ) / i Ai j =1 ( Ai Fij ) 1

que pode ser representado


por um anlogo elctrico do tipo:
Metodologia de anlise para recintos fechados
 Aplicar Eq. (4) a cada superfcie para a qual o fluxo til de radiao qi
connecido.
 Aplicar Eq. (5) a cada uma das restantes superfcies para a qual a temperatura Ti ,
e, portanto, Ebi ,, conhecida.

 Calcular todos os factores de forma que aparecem nas equaes

 Resolver o sistema de N equaes para as radiosidades (incgnitas) J1 , J 2 ,...., J N .

 Usar Eq. (3) para determinar qi para cada superfcie onde se conhece
Ti e para determinar Ti para cada superfcie onde se conhece qi.

Tratamento de superfcie virtual correspondente a abertura de rea Ai , atravs da


as superfcies interiores de um recinto fechado trocam radiao com a envolvente
(de grandes dimenses) temperatura Tsur :

 Aproximar a abertura a um corpo negro de rea, Ai , temperatura, Ti = Tsur ,


e propriedades, i = i = 1 .
Recintos fechados com duas superfcies
Recinto mais simples para o qual a troca de calor por radiao se d
exclusivamente entre duas superfcies e em que uma expresso para a
troca de calor por radiao pode ser determinada directamente atravs do
anlogo elctrico.

(T14 T24 )
q1 = q2 = q12 =
1 1 1 2
+ 1 +
1 A1 A1F12 2 A2
A tabela 13.3 apresenta resultados para alguns casos especiais. Por exemplo:

 Placas paralelas e infinitas

A1 = A2 A A1 (T14 T24 )
q12 =
F12 = 1 1 + 1 1
1 2

 Pequena superfcie plana/convexa rodeada por uma superfcie muito maior

As
=0
Asur
(
q = As s Ts4 Tsur
4
)
Fs , sur = 1
Escudo de radiao
=
Superfcie com reflectividade elevada (baixo = ) colocada entre duas
superfcies cuja troca de calor por radiao se pretende reduzir

Considere um nico escudo de radiao num reconto fechado, tal como o caso de
duas placas planas paralelas e infinitas.

Note que, embora raramente, a emisso pode ser diferente para as duas superfcies do
escudo de radiao.
Anlogo elctrico

(T14 T24 )
q12 = q1 = q2 =
1 1 1 3,1 1 3,2 1 2
+ 1 + + + 1 +
1 A1 A1F13 3,1 A3 3,2 A3 A3 F32 2 A2

O resultado anterior pode ser facimente estendido para ter em conta mlitplos
escudos de radiao e aplicado a cilindros longos e concntricos, esferas
concntricas e placas longas e paralelas.
A superficie re-radiante

Uma idealizao para a qual: GR = J R . Portanto, qR = 0 and J R = EbR .

Corresponde a superfcies que so bem isoladas de um lado e para as quais


a conveco desprezvel do lado oposto (radiante).

Recinto fechado com trs superfcies sendo uma delas re-radiante :

(T14 T24 )
q1 = q2 =
1 1 1 1 2
+ +
1 A1 A1F12 + (1 / A1F1R ) + (1 / A2 F2 R ) 1 2 A2
A superficie re-radiante

Temperatura da superfcre re-radiante TR pode ser determinada a partir do conhecimento


da sua radiosidade J R .

J1 J R JR J2
=
(1 / A1F1R ) (1 / A2 F2 R )

1/ 4
JR
TR =

Problem 13.88: Power requirement for a cylindrical furnace with two
reradiating surfaces and an opening to large surroundings.

KNOWN: Cylindrical furnace of diameter D = 90 mm and overall length L = 180 mm.


Heating elements maintain the refractory lining ( = 0.8) of section (1), L1 = 135 mm, at T1 =
800C. The bottom (2) and upper (3) sections are refractory lined, but are insulated. Furnace
operates in a spacecraft vacuum environment.

FIND: Power required to maintain the furnace operating conditions with the surroundings at
23C.
SCHEMATIC:

ASSUMPTIONS: (1) All surfaces are diffuse gray, and (2) Uniform radiosity over the
sections 1, 2, and 3.

ANALYSIS: By defining the furnace opening as the hypothetical area A4, the furnace can be
represented as a four-surface enclosure.

The power required to maintain A1 at T1 is q1, the net radiation leaving A1.

To obtain q1, we must determine the radiosity at each surface by simultaneously solving
radiation energy balance equations of the form
E bi Ji N J J
j j
q = = (1,2)
i (1 i ) / i Ai j =1 1/ Ai Fij
However, since 4 = 1, J4 = Eb4, and only three energy balances are needed for A1, A2, and A3.
E b1 J1 J1 J 2 J J J J4
A1: = + 1 3 + 1 (3)
(1 1 ) / 1A1 1/ A1F12 1/ A1F13 1/ A1F14
J J J J3 J J4
A2: 0= 2 1 + 2 + 2 (4)
1/ A 2 F21 1/ A 2 F23 1/ A 2 F24
J3 J1 J J2 J J4
A3: 0= + 3 + 3 (5)
1/ A3 F31 1/ A3 F32 1/ A3 F34
where q2 = q3 = 0 since the surfaces are insulated (adiabatic) and hence reradiating.

From knowledge of J1, q1 can be determined using Eq. (1).

Of the N2 = 42 = 16 view factors, N(N 1)/2 = 6 must be independently evaluated, while the
remaining can be determined by the summation rule and appropriate reciprocity relations. The
six independently determined Fij are:

By inspection: (1) F22 = 0 (2) F44 = 0


Coaxial parallel disks: From Table 13.2,

1/ 2
F24 = 0.5 S S2 4 ( r4 / r2 )
2
(3)
= 0.05573

where
2 2
1 + R4 1 + 0.250
S = 1+ =1+ = 18.00 R 2 = r2 / L = 45 / 180 = 0.250 R 4 = r4 / L = 0.250
2 2
R2 0.250

Enclosure 1-2-2: From the summation rule for A2,

(4) F21 = 1 F22 = 1 0.09167 = 0.9083


where F22 can be evaluated from the coaxial parallel disk relation, Table 13.2, with R2 = r2/L1 =
45/135 = 0.333, R2 = r2/L1 = 0.333, and S = 11.00.

From the summation rule for A1,


(5) F11 = 1 F12 F12 = 1 0.1514 0.1514 = 0.6972
From symmetry F12 = F12 and using reciprocity
F12 = A 2 F21 / A1 = [ ( 0.090m )( 2 / 4 )] 0.9083 / 0.090m 0.135m = 0.1514

Enclosure 2 -3-4: From the summation rule for A4,


(6) F43 = 1 F42 F44 = 1 0.3820 0 = 0.6180
where F44 = 0 and using the coaxial parallel disk relation from Table 13.2, F42 =0.3820 with R4 =
r4/L2 = 45/45 = 1, R2 = r2/L2 = 1, and S = 3.
The View Factors: Using summation rules and appropriate reciprocity relations, the remaining 10 view
factors can be evaluated. Written in matrix form, the Fij are
0.6972* 0.1514 0.09704 0.05438
0.9083* 0* 0.03597 0.05573*
0.2911 0.01798 0.3819 0.3090
0.3262 0.05573 0.6180* 0*
The Fij shown with an asterisk were independently determined.

From knowledge of the relevant view factors, the energy balances, Eqs. (3, 4, 5), can be solved
simultaneously to obtain the radiosities,
J1 = 73, 084 W / m 2 J 2 = 67, 723 W / m 2 J 3 = 36, 609 W / m 2

The net heat rate leaving A1 can be evaluated using Eq. (1) written as
E b1 J1 ( 75,159 73, 084 ) W / m 2
q1 = = = 317 W <
(1 1 ) / 1A1 (1 0.8 ) / 0.8 0.03817 m 2
where Eb1 = T14 = (800 + 273K)4 = 75,159 W/m2 and A1 = DL1 = 0.090m 0.135m =
0.03817 m2.

COMMENTS: Recognize the importance of defining the furnace opening as the hypothetical
area A4 which completes the four-surface enclosure representing the furnace. The temperature
of A4 is that of the surroundings and its emissivity is unity since it absorbs all radiation incident
on it.
Problem 13.93: Assessment of ceiling radiative properties for an ice rink
in terms of ability to maintain surface temperature above
the dewpoint.

KNOWN: Ice rink with prescribed ice, rink air, wall, ceiling and outdoor air conditions.

FIND: (a) Temperature of the ceiling, Tc, for an emissivity of 0.05 (highly reflective panels) or
0.94 (painted panels); determine whether condensation will occur for either or both ceiling
panel types if the relative humidity of the rink air is 70%, and (b) Calculate and plot the ceiling
temperature as a function of ceiling insulation thickness for 0.1 t 1 m; identify conditions
for which condensation will occur on the ceiling.
SCHEMATIC:

ASSUMPTIONS: (1) Rink comprised of the ice, walls and ceiling approximates a three-
surface, diffuse-gray enclosure, (2) Surfaces have uniform radiosities, (3) Ice surface and walls
are black, (4) Panels are diffuse-gray, and (5) Thermal resistance for convection on the outdoor
side of the ceiling is negligible compared to the conduction resistance of the ceiling insulation.
PROPERTIES: Psychometric chart (Atmospheric pressure; dry bulb temperature, Tdb = T,i
= 15C; relative humidity, RH = 70%): Dew point temperature, Tdp = 9.4C.

ANALYSIS: Applying an energy balance to the inner surface of the ceiling and treating all
heat rates as energy outflows,
E& in E& out = 0
q o q conv,c q rad,c = 0 (1)
where the rate equations for each process are
( )
q o = Tc T,o / R cond R cond = t / kA c (2,3)

(
q conv,c = h i A c Tc T,i ) (4)
q rad,c = E b ( Tc ) Ac A w Fwc E b ( Tw ) Ai Fic E b ( Ti ) (5)

Since the ceiling panels are diffuse-gray, = .


From Table 13.2 for parallel, coaxial disks
Fic = 0.672

From the summation rule applied to the ice (i) and the reciprocity rule,
Fic + Fiw = 1 Fiw = Fcw (symmetry)
Fcw = 1 Fic
Fwc = ( A c / A w ) Fcw = ( A c / A w ) (1 Fic ) = 0.410
where Ac = D2/4 and Aw = DL.
Using the foregoing energy balance, Eq. (1), and the rate equations, Eqs. (2-5), the ceiling
temperature is calculated using radiative properties for the two panel types,
Ceiling panel Tc (C)
Reflective 0.05 14.0
Paint 0.94 8.6 Tc < Tdp <

. Condensation will occur on the painted panel since Tc < Tdp.

b) Applying
(b) Applyingthe
theforegoing
foregoingmodel
modelforfor tt1.01.0
0.10.1 m the
m, the following
following result
result is obtained
is obtained

15
Ceiling temperature, Tc (C)

10

5
0 0.2 0.4 0.6 0.8 1
Ceiling insulation thickness, t (m)

Painted ceiling, epsc = 0.94


Reflective panel, epsc = 0.05
For the reflective panel ( = 0.05), the ceiling surface temperature is considerably above the
dew point. Therefore, condensation will not occur for the range of insulation thicknesses. For
the painted panel ( = 0.94), the ceiling surface temperature is always below the dew point, and
condensation occurs for the range of insulation thicknesses.

COMMENTS: From the analysis, recognize that radiative exchange between the ice and the
ceiling has the dominant effect on the ceiling temperature. With the reflective panel, the rate is
reduced nearly 20-fold relative to that for the painted panel. With the painted panel ceiling,
condensation will occur for most of the conditions likely to exist in the rink.