1. Trang chủ
  2. » Kỹ Thuật - Công Nghệ

Heat Transfer Handbook part 4 ppt

10 313 0
Tài liệu đã được kiểm tra trùng lặp

Đang tải... (xem toàn văn)

THÔNG TIN TÀI LIỆU

Thông tin cơ bản

Định dạng
Số trang 10
Dung lượng 166,54 KB

Các công cụ chuyển đổi và chỉnh sửa cho tài liệu này

Nội dung

Equation 1.64 is the statement of the conservation of momentum principle.. Note that the conservation of momentum principle is stated in terms of the proper-ties of particles and not in

Trang 1

20 BASIC CONCEPTS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[20],(20)

Lines: 940 to 1001

———

0.99356pt PgVar

———

Normal Page PgEnds: TEX [20],(20)

F= m d ˆV

dt =

d(m ˆV)

wheremV is the momentum Equation (1.64) is the statement of the conservation of

momentum principle.

Note that the conservation of momentum principle is stated in terms of the proper-ties of particles and not in terms of the properproper-ties of a field To derive the momentum theorem, a region in a fluid confined by the control surfaceS1, shown in Fig 1.7, is employed The surfaceS1contains a definite and fixed number of particles at timet1

At timet2, these particles will have moved to a region bounded by the control surface

S2, which is shown as a dashed curve to distinguish it fromS1 The control surfacesS1andS2enclose three separate and distinct regions, desig-nated bya,b, and c Let the momentum in the three regions be P a , P b, and Pc, re-spectively At timet1the particles within surfaceS1will possess momentum Pa+Pb1

At timet2these particles will have momentum, Pb2+ Pcbecause they have moved into the region enclosed by surfaceS2 Hence the momentum change during the time intervalt2− t1may be described by

(P b2+ Pc ) − (P b1+ Pa ) = (P b2− Pb1 ) + (P c− Pa )

and the time rate of change of momentum will be

lim

t2→t1



Pb2− Pb1

t2− t1

+Pc− Pa

t2− t1



(1.65)

Ast2 approachest1 as a limit, the control surfaceS2 will coincide withS1 The first term in eq (1.65) is therefore the time rate of change of momentum of the fluid contained within region 1, R1, contained withinS1 This may be written as the integral overR1 Because the mass of fluid contained inR1is



R1

ρ dR1

the time rate of change of momentum of the fluid contained within region 1 will be

∂t



R1

ρ ˆV dR1

The second term in eq (1.65) is the momentum efflux through the control surface

S1 If the flux in the outward direction is taken as positive, this efflux can be expressed

by the integral



S1

ρ ˆV ˆV n dS1 whereV nis the component of velocity normal toS1

Trang 2

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[21],(21)

Lines: 1001 to 1042

———

0.42567pt PgVar

———

Normal Page PgEnds: TEX [21],(21)

c b a

Figure 1.7 Regions bounded by control surfaces used for the development of the momentum theorem

The conservation of momentum principle then becomes

F = m d ˆV

dt =

∂t



R1

ρ ˆV dR1+



S1

ρ ˆV ˆV n dS1 (1.66)

or, by rearrangement of terms,

∂t



R1

ρ ˆV x dR1= F x



S1

ρ ˆV x ˆV n dS1 (1.67a)

∂t



R1

ρ ˆV y dR1= F y



S1

ρ ˆV y ˆV n dS1 (1.67b)

∂t



R1

ρ ˆV z dR1 = F z



S1

ρ ˆV x ˆV n dS1 (1.67c)

in the three rectangular coordinate directions

The foregoing development leads to the statement of the momentum theorem: The

time rate of increase of momentum of a fluid within a fixed control volumeR will

be equal to the rate at which momentum flows intoR through its confining surface

S, plus the net force acting on the fluid within R When the flow is incompressible, the viscosity is constant, and the flow is laminar, the Navier–Stokes equations result.

In Cartesian coordinates, withF x,F y, andF ztaken as the components of the body force per unit volume, the Navier–Stokes equations are

ρ



∂ ˆV x

∂t + ˆV x

∂ ˆV x

∂x + ˆV y

∂ ˆV x

∂y + ˆV z

∂ ˆV z

∂z



Trang 3

22 BASIC CONCEPTS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[22],(22)

Lines: 1042 to 1077

———

1.68011pt PgVar

———

Long Page

* PgEnds: Eject

[22],(22)

= −∂P ∂x + µ



2ˆV x

∂x2 +2ˆV x

∂y2 +2ˆV x

∂z2



+ F x (1.68a)

ρ



∂ ˆV y

∂t + ˆV x

∂ ˆV y

∂x + ˆV y

∂ ˆV y

∂y + ˆV z

∂ ˆV y

∂z



= −∂P

∂y + µ



2ˆV y

∂x2 +2ˆV y

∂y2 +2ˆV y

∂z2



+ F y (1.68b)

ρ



∂ ˆV z

∂t + ˆV z

∂ ˆV x

∂x + ˆV y

∂ ˆV z

∂y + ˆV z

∂ ˆV z

∂z



= −∂P

∂z + µ



2ˆV z

∂x2 +2ˆV z

∂y2 +2 ˆV z

∂z2



+ F z (1.68c)

In cylindrical coordinates withF r Fθ, andF ztaken as the components of the body

force per unit volume, the Navier–Stokes equations are

ρ



∂ ˆV r

∂t + ˆV r

∂ ˆV r

∂r +

ˆVθ

r

∂ ˆV r

∂θ

ˆV2

θ

r + ˆV z

∂ ˆV r

∂z



= −∂P

∂r

+ µ



2ˆV r

∂r2 +1r ∂ ˆV ∂r rˆV r

r2 +r122ˆV r

∂θ2 −r22 ˆVθ

∂θ+

2ˆV r

∂z2



+ F r (1.69a)

ρ



∂ ˆVθ

∂t + ˆV r

∂ ˆVθ

∂r +

ˆVθ

r

∂ ˆVθ

∂θ

ˆV r ˆVθ

r + ˆV z

∂ ˆVθ

∂z



= −1

r

∂P

∂θ

+ µ



2ˆVθ

∂r2 +1r ∂ ˆV ∂rθ− ˆVθ

r2 +r122 ˆVθ

∂θ2 +r22∂ ˆV ∂θ r +2 ˆVθ

∂z2



+ Fθ (1.69b)

ρ



∂ ˆV z

∂t + ˆV r

∂ ˆV z

∂r +

ˆVθ

r

∂ ˆV z

∂θ + ˆV z

∂ ˆV z

∂z



= −∂P

dz

+ µ



2ˆV z

∂r2 +1

r

∂ ˆV z

∂r +

1

r2

2ˆV z

∂θ2 +2ˆV z

∂z2



Finally, in spherical coordinates withF r Fθ, andFφ taken as the components of the body force per unit volume and with

D

DT =

dt + ˆV r

∂r +

ˆVφ

r

∂φ +

ˆVθ

r sin φ

∂θ

∇2= 1

r2

∂r



r2

∂r

 + 1

r2sinφ

∂φ



sinφ

∂φ

 + 1

r2sin2φ

2

∂θ2

Trang 4

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[23],(23)

Lines: 1077 to 1154

———

2.01216pt PgVar

———

Long Page

* PgEnds: Eject

[23],(23)

the Navier–Stokes equations are

ρ



D ˆV r

Dt

ˆV2

φ + ˆV2

θ

r



= −∂P ∂r + µ



∇2ˆV r−2 ˆV r

r2 − 2

r2

∂ ˆVφ

∂φ

2 ˆVθcotφ

r2 − 2

r2sinφ

∂ ˆVθ

∂θ



+ F r (1.70a)

ρ



D ˆVφ

Dt

ˆV r ˆVφ

r

ˆV2

θ cotφ

r



= −1

r

∂P

∂φ

+ µ



∇2ˆVφ+r22∂ ˆV ∂φ rˆVφ

r2sinφ−

2 cosφ

r2sin2φ

∂ ˆVθ

∂θ



+ Fφ (1.70b)

ρ



D ˆVθ

Dt +

ˆVθˆV r

r +

ˆVφ ˆVθcotφ

r



= − ˆVθ

r sin φ

∂P

∂θ

+ µ



∇2ˆVθ− ˆVθ

r2sin2φ+

2

r2sinφ

∂ ˆV r

∂θ +

2 cosφ

r2sin2φ

∂ ˆVθ

∂θ



+ Fθ (1.70c)

In Fig 1.8, an imaginary two-dimensional control volume of finite size∆x ∆y with

flow velocity ˆV = ex ˆV x+ ey ˆV y, heat flux q= ex q

x+ ey q

y, specific internal energy

u, and rate of internal heat generation q, the first law of thermodynamics requires

that

 rate of energy

accumulation within the control volume



=

 net transfer

of energy by fluid flow

 +

 net heat

transfer by conduction



×

 rate of

internal heat generation



net work transfer from the control volume to the environment

 (1.71)

Four of the five terms indicated do not involve work transfer from the control volume

to the environment

• The rate of energy accumulated in the control volume is

∆x ∆y ∂t ∂ (ρu) (1.72a)

Trang 5

24 BASIC CONCEPTS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[24],(24)

Lines: 1154 to 1178

———

2.91211pt PgVar

———

Long Page

* PgEnds: Eject

[24],(24)

( )ρe

t ⌬ ⌬x y

y

qxy q x⵮⌬ ⌬y

qyx

q⬙ ⫹x ∂q ∂x x x⬙ ⌬ ⌬y

(

(

)

)

ρV uˆy∂y V u y ∂ (ρˆy )⌬ ⌬ x

ρV uˆx∂y V u x ∂ (ρˆx )⌬ ⌬ y

ρ ˆV u y x 

ρ ˆV u x y 

Figure 1.8 First law of thermodynamics applied to an imaginary control volume in two-dimensional flow

• The net transfer of energy by fluid flow is

− (∆x ∆y)



∂x



ρ ˆV x u+∂y ∂ ρ ˆV y u (1.72b)

• The net heat transfer by conduction is

− (∆x ∆y)



∂q

x

∂x +

∂q

y

∂y



(1.72c)

Trang 6

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[25],(25)

Lines: 1178 to 1196

———

3.754pt PgVar

———

Long Page PgEnds: TEX [25],(25)

• The rate of internal heat generation is

The origin of the term involving the net work transferred from the control volume

to the environment is shown in Fig 1.9, where the normal and tangential stresses are sketched For example, the work done per unit time by the normal stressσxon the left

ρV Vˆ ˆx y∂y V V ∂ (ρˆ ˆx y) ⌬ ⌬y x

∂t (ρˆV x) ⌬ ⌬y x

ρˆ2⫹ ( )ρ∂xˆ2 ⌬x⌬y

␶ ⫹xy ∂ ∂yxy ⌬ ⌬y x

xy ⌬x

xy

ρˆV x2⌬y

ˆ ˆ

V V x yx

␴ ⫹x ∂ ∂xx ⌬ ⌬x y

X x y⌬ ⌬

(

(

)

)

y

x

Figure 1.9 Force balance in thex direction of an imaginary control volume in two-dimensional

flow

Trang 7

26 BASIC CONCEPTS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[26],(26)

Lines: 1196 to 1263

———

3.16624pt PgVar

———

Long Page

* PgEnds: Eject

[26],(26)

side of the∆x ∆y element is negative and equal to the force acting on the boundary

σy, multiplied by the boundary displacement per unit time ˆV x This yields− ˆV xσxy.

Similarly, the work transfer associated with normal stresses acting on the right side

of the element is positive and equal to



σx+∂σ ∂x x ∆x

 

ˆV x+∂ ˆV ∂x x ∆x



∆y

The net work transfer rate due to these two contributions is



σx ∂ ˆV x

∂x + ˆV x

∂σ x

∂x



(∆x ∆y)

Three more work transfer rates can be calculated in the same way by examining the effect of the remaining three stresses,τxyin thex direction and σ yandτyxin the

y direction This gives

(∆x ∆y)



σx ∂ ˆV x

∂x − τxy

∂ ˆV x

∂y + σy

∂ ˆV y

∂y − τyx

∂ ˆV y

∂x



+ (∆x ∆y)



ˆV x ∂σ x

∂x − ˆV x

∂τ xy

∂y + ˆV y

∂σ y

∂y − ˆV x

∂τ yx

∂x



where the eight terms have been separated into two groups It can be shown that the second group reduces to

− ρDt D ˆV2

x + ˆV2

y

2 which represents the change of kinetic energy of the fluid in the control volume This change may be considered negligible relative to the internal energy change,∂(ρu)/∂t,

so that the work transfer becomes

(∆x ∆y)



σx ∂ ˆV x

∂x − τxy

∂ ˆV x

∂y + σy

∂ ˆV y

∂y − τyx

∂ ˆV y

∂x



(1.72e)

The stressesσxandτxycan be related to the flow field via the constitutive relations given by Rohsenow and Choi (1961):

σx = P − 2µ ∂ ˆV x

∂x +

2



∂ ˆV x

∂x +

∂ ˆV y

∂y



(1.73a) and

τxy = µ



∂ ˆV x

∂y +

∂ ˆV y

∂x



(1.73b)

Trang 8

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[27],(27)

Lines: 1263 to 1328

———

0.00224pt PgVar

———

Long Page

* PgEnds: Eject

[27],(27)

The bookkeeping required by eq (1.71) dictates the assembly of eqs (1.72), using eqs (1.73), into

ρDu

Dt + u



Dp

Dt + ρ∇ · ˆV



= −∇ · q+ q− P∇ · ˆV + µΦ (1.74)

whereµ is the dynamic viscosity of the fluid and Φ is the viscous dissipation function,

which is detailed subsequently in rectangular, cylindrical, and spherical coordinates

However, eq (1.53) shows that the term in parentheses on the left-hand side of eq

(1.74) is equal to zero, so that eq (1.74) reduces to

ρDu

Dt = − ∇ · q+ q− P∇ · ˆV + µΦ (1.75)

In the special case where the flow can be modeled as incompressible and two-dimensional, the viscous dissipation function reduces to

Φ = 2



∂ ˆV x

∂x

2

+



∂ ˆV y

∂y

2

 +



∂ ˆV x

∂y +

∂ ˆV y

∂x

2

(1.76)

To express eq (1.74) in terms of enthalpy, the definition

h = u + P v = u + P

ρ

is invoked Hence,

Dh

Dt =

Du

Dt +

1

ρ

DP

Dt

P

ρ2

Moreover, the heat fluxesq

x andq

y can be expressed in terms of local temperature

gradients through use of Fourier’s law:

Thus, the combination of eqs (1.74), (1.77), and (1.78) results in

ρDh

Dt = ∇ · (k ∇T ) + q+

DP

Dt + µΦ −

P

ρ



Dt + ρ ∇ · ˆV



(1.79)

Here, too, eq (1.53) points out that the terms in parentheses in eq (1.79) are equal to zero, so that eq (1.75) reduces to

ρDh

Dt = ∇ · (k ∇T ) + q+

DP

Bejan (1995) points out that the change in specific enthalpy for a single-phase fluid

is given by

Trang 9

28 BASIC CONCEPTS

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[28],(28)

Lines: 1328 to 1411

———

7.50035pt PgVar

———

Long Page PgEnds: TEX [28],(28)

dh = T ds + v dP = T ds + dP

whereT is the absolute temperature and ds is the specific entrophy change:

ds =



ds dT



P dT +



ds dP



The last of the Maxwell relations given by Bejan (1997)

∂s

∂P



T = −

∂(1/ρ)

∂T



P = ρ12

∂ρ

∂T



P = −βρ (1.83) whereβ is the volumetric coefficient of thermal expansion,

β = −1 ρ



∂ρ

∂T



P

(1.84)

Withc p taken as the specific heat at constant pressures, it can be shown that

ds

dT



P = c p

and eqs (1.81) through (1.85) can be combined,

dh = c p dT + 1

ρ(1 − βT ) dP (1.86)

so that the left-hand side of eq (1.80) can be written as

ρDh

DT = ρc p

DT

Dt + (1 − βT )

DP

Thus, the temperature formulation of the first law of thermodynamics is

ρc p DT

Dt = ∇ · (k ∇T ) + q+ βT

DT

with the special forms for the ideal gas whereβ = 1/T ,

ρc p DT

Dt = ∇ · (k ∇T ) + q+

DP

and for an incompressible fluid whereβ = 0,

ρc p DT

Dt = ∇ · (k ∇T ) + q+ µΦ (1.89b)

Most convection problems concern an even simpler model where the fluid has constant thermal conductivityk, neglible viscous dissipation Φ, zero internal heat

Trang 10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45

[29],(29)

Lines: 1411 to 1462

———

9.43025pt PgVar

———

Long Page

* PgEnds: Eject

[29],(29)

generation (q = 0), and a negligible compressibility effect, βT (DP /Dt) ≈ 0 The

energy equation for this model is simply

ρc p DT

Dt = k ∇2T (1.90)

or, in the rectangular coordinate system,

ρc p

∂T

∂t + ˆV x

∂T

∂x + ˆV y

∂T

∂y + ˆV z

∂T

∂z



= k

2T

∂x2 +2T

∂y2 +2T

∂z2



(1.91)

in the cylindrical coordinate system,

ρc p



∂T

∂t + ˆV r

∂T

∂r +

ˆVθ

r

∂T

∂θ + ˆV z

∂T

∂z



= k



1

r

∂r



r ∂T ∂r

 +r122T

∂θ2 +∂ ∂z2T2



(1.92) and in the spherical coordinate system,

ρc p



∂T

∂t + ˆV r

∂T

∂r +

ˆVφ

r

∂T

∂φ +

ˆVθ

r sin φ

∂T

∂θ



= k



1

r2

∂r



r2∂T

∂r

 +r2 1

sinφ

∂φ



sinφ∂T ∂φ

 + 1

sin2φ

2T

∂θ2



(1.93)

If the fluid can be modeled as incompressible then, as in eq (1.89b), the specific heat at constant pressure c p is replaced by c And when dealing with extremely

viscous flows, the model is improved by taking into account the internal heating due

to viscous dissipation,

ρc p DT

Dt = k ∇2T + µΦ (1.94)

In the rectangular coordinate system, the viscous dissipation can be expressed as

Φ = 2



∂ ˆV x

∂x

2

+



∂ ˆV y

∂y

2

+



∂ ˆV z

∂z

2

+



∂ ˆV x

∂y +

∂ ˆV y

∂x

2

+



∂ ˆV y

∂z +

∂ ˆV z

∂y

2

+



∂ ˆV z

∂x +

∂ ˆV x

∂z

2

−2

3



∂ ˆV x

∂x +

∂ ˆV y

∂y +

∂ ˆV z

∂z

2

(1.95)

Ngày đăng: 05/07/2014, 16:20

TỪ KHÓA LIÊN QUAN