Shell Energy Balance and Temperature Distributions in Solids and Laminar Flows

Download as doc, pdf, or txt
Download as doc, pdf, or txt
You are on page 1of 13

Chapter 10

Shell Energy Balance and Temperature Distributions in Solids and Laminar Flows

Conservation law:
Rate of ( ) in – rate of ( ) out + rate of ( ) generation = rate of ( ) accumulation

Conservation of mass:

Continuity equation:

Conservation of momentum:

Cauchy’s Equation:

Navier-Stoke’s Equation:
(for incompressible Newtonian fluid)

Conservation of energy:

Energy equation:

Textbook (10.1-1) is too complicate

§ 10.2 Heat Conduction with a Electrical Heat Source Se (W/m3)


Find the steady state temperature distribution of an electrical wire with internal heat generation
(Fig.10.2-1).
 By shell energy balance: In cylindrical coordinate, take a shell (C.V.) with thickness r, radius r and
length L, then apply conservation of energy to the CV to obtain

Divided the equation by 2rrL, and let r0, we have


, or .

C1 has to be zero.

So , Hence

After applying BC at r = R, T = T0 we have

QED

 By simplifying eqn. of energy and note that in cylindrical

coordinate. Since , we have =-

 Three kinds of BC in heat transfer problems: (i ) Constant TW, (ii) constant qw, and (iii) Newton’s
Law of Cooling applies at the solid and fluid boundary.
§ 10.3 Steady State Heat conduction with a Nuclear Heat Source (Fig.10.3-1)
 Assume that the heat source can be approximated by (1)

 Make shell energy balance for the sphere of fissionable, we have

, or (2)

For aluminum cladding, we have (3)

Integration of (2) and (3) yield …..(4) and (5)

 From BC1: at r = 0, ; and BC2: at r = RF

We have = (6)

(7).

Integrating (6) and (7) to obtain:

and

2 and 2 can be found from BC3: at r = RF, TF = TC; and BC4 at r = RC, TC = T0. The final
temperature distributions are given by Eqns (10.3-20) and (10.3-21).

§ 10.4 Heat Conduction with a Viscous Heat Source (Fig.10.4-1 and 10.4-2)

 Assume1, so that curvature effect can be neglected.


 From energy eqn. (1)

where  (W/m3) arises from viscous dissipation. We know from previous analysis that work flux
(J/m2.s = W/m2) = is a vector and .
What is  in this case?

Since v1 = v2 = 0, v3 = vz(x); we have since .

Hence . Since DT/Dt = 0 in this case since vx = vy = 0 and

Eqn (1) reduces to (2)

 If we start from shell energy balance:


(Convection in) + (conduction in) – (convection out) – (conduction out) + generation = 0

or (3)
the same as before.
On p.341, you will see either or or : .
The relations between those 3 are given by A.4-29 = : + .
Integrating of (3) with vz = vb(x/b) and BC: at x = 0, T = T0 and at x = b, T = Tb gives

where = Brinkman number which is a measure of the importance of viscous

dissipation.
Q: Where is the maximum temperature?

§ 10.5 Heat Conduction with a Chemical Heat Source (Fig.10.5-1)

Zone I: Entrance zone packed with non-catalytic spheres


Zone II: Reaction zone packed with catalytic spheres, from z = 0 to z = L
Zone III: Exit zone packed with non-catalytic spheres

 Energy balance over CV z.A, assume vz = constant and  = constant

Or ;

this can be obtained more easily from .

Equations (10.5-6), (10.5-7) and (10.5-8) for zone I, II and III, respectively.
 6 BCs, from (10.5-9) to (10.5-14), are needed since there are three 2nd order ODE to be solved.
Important concept: Continue of temperature and heat flux at interface I, II and II, III.

§ 10.6 Heat Conduction through Composite Walls (Fig.10.6-1, 10.6-2)

1. Composite plane walls:,

2. Composite cylinders: where Aj,lm is the log-mean of adjacent

layers

3. Composite spheres: where Aj,gm is the geometric-mean of

adjacent layers

§ 10.7 Heat Conduction in Cooling Fins (Fig.10.7-1)


 For straight fin, assume (1) T = T(z) only, (2) No heat loss from the end or edges of the fin, (3) heat
loss from the surface is q = h(T – Ta).
 The shell energy balance across CV = z(2B)W is
, hence or

(1).

 The pertinent BCs are (1) at z = 0, T = Tw; and (2) at z = L, dT/dz = 0 (or more accurately, -kdT/dz
= h(T – Ta)). The solution of (1) with the pertinent BCs is

where ,  = z/L, and .

Note where Bi = Biot number = (resistance to conduction)/(resistance to

convection)
 Define fin efficiency  as  = (actual rate of heat loss from the fin)/(ideal rate of heat loss from the
fin). Actual heat loss =

which approaches

2hWL(Tw –Ta) as N0, i.e. k (ideal case!). Note that in ideal case the whole fin will have a
temperature of Tw. Hence  = tanhN/N. QED
 For circular fin with thickness  and radii R and R, shell energy balance results in
, i.e. or
(1).

This is put in dimensionless form as (2)

where x = r/R, , and .

Eqn (2) can be put into x2” +x’-x2N2 = 0


which is a modified Bessel Equation, the solution is q = c1I0(Nx) + c2K0(Nx). I0 and K0 are 0th order
1st kind modified Bessel function and 0th order 2nd kind modified Bessel function, respectively.

§ 10.8 Forced Convection (Fig.10.8-2)

 Energy balance over CV with thickness r and z to obtain


+ +
(1)

Where in this case.

Eqn(1) can be reduced to or

(2).
Again, this can be obtained directly from . After neglecting viscous

dissipation and conduction in z-direction, we have (3)

 BCs as: (i) at r = 0, T = finite; (ii) at r = R, k∂T/∂r = q0 = const.; (iii) z = 0, T = T0.

Eqn(3) with BCs can be transformed into (4)

with BCs as (i)  = 0,  = finite; (ii)  = 1, ∂/∂ = 1; and (iii)  = 0,  = 0.


The complete solution of Eqn(4) will be given in Example 12.2-1. Asymptotic solution for small 
will be given in Example 12.2-2.
We will give asymptotic solution for large  here.
For large  is linear in , i.e.  = c0 + () (5)
and BC (iii) is replaced by

or in dimensionless form …. (iv)

where

,  = r/R and  =z/(vmaxR2) = (z/R)()(/Dvav) =(z/R)Pr -1Re -1 =

Substitute (5) into (4) we have

Integrating twice to obtain  = c02/4 –c04/16 + c1ln + c2.

From BC (i), c1 = 0; from BC (ii) , hence c0 = 4.

From BC (iv) or

so that c2 = -7/24

What is average temperature?

Cross-sectional area average temperature

Or
=

Mixing –cup temperature Tb is defined as

So that

Or (6)

Since we have

(7)

(7)-(6) yields , so that we have Nu = hD/k = 48/11 = 4.364.

This is the Nusselt number at large  for constant wall flux. QED

Alternative approach (The materials presented here is from “Convective Heat and Mass Transfer” by
William M. Kays and Michael E. Crawford)

For fully developed temperature profile (FDTP) is independent of z.

Hence for FDTP, Nu = hD/k = const. for both const qw and const. Tw.

Since for FDTP, we have ,

or .

For const qw, qw =h(Tw – Tb) = const, we have so energy eqn becomes
For const Tw, so energy eqn becomes

Case 1. Const qw

Integrate twice with BCs (i) at r = 0, ∂T/∂r = 0, and (ii) at r = R, T = Tw, we have

, then .

Hence qw = h(Tw – Tb) = (8)

Make energy balance over CV of thickness dz to obtain qw.2Rdz = vavR2CPdTb, or

(9)

Substitute (9) into (8) to get

Case 2 Const Tw (By iteration)

…(10). Note in (10) is an unknown function of r.

(a) Assume , substitute T(0) into the RHS of (10) and obtain

T(1) in the LHS, then calculate Nu(1).


(b) Substitute T(1) into the RHS of (10) and obtain T(2) in the LHS, then calculate Nu(2).
(c) Etc.

NuT = 3.66

§ 10.9 Free Convection (Fig.10.9-1)


 Assume (i) T= TH - TC is small; and (ii) Tall plates, end effects can be neglected.
 Energy balance over CV gives – dqy/dy = 0, or kd2T/dy2 = 0. Integrating of this eqn subject to
BCs: at y = B, T = TC and at y = - B, T = TH gives (1)

where and .

 From Navier-Stokes eqn ( is function of temperature): (2)

the z-component of (2) is (3)

Expand  by Taylor series and we have , or

where , So (3) becomes

(4)

Substitute (1) into (4) to get …. (5)

Eqn(5) is to be solved with BCs that vz = 0 at y = B. The solution is

(6)

If the flow is in an enclosure, then up flow = down flow, and . By noting that the
integration of odd function from –a to +a is zero, we obtain so that

(7)

Note that

Gr is called Grashof number.

Eqn(7) can be put into QED

You might also like