CBE 120C – Mass Transfer

Spring 2023, Homework 1 Solution

PROBLEM 1

a. 𝑦 ′′ + 𝑦 = 0.001𝑥 2 BCs: 𝑦(0) = 0, 𝑦 ′ (50) = 1.5

This is a second order non-homogeneous constant coefficient ODE. The solution is the sum of the
complementary and particular solutions. For the complementary solution, since it is a constant
coefficient ODE, write the characteristic equation:

𝑟 2 + 1 = 0 ⟹ 𝑟 = ±𝑖
𝑦𝑐 = 𝑐1 cos(𝑥) + 𝑐2 sin⁡(𝑥)
For the particular solution, choose the function based on the form of the inhomogeneity
𝑦𝑝 = 𝑐3 𝑥 2 + 𝑐4 𝑥 + 𝑐5
Substitute this into the differential equation, set coefficients of like terms equal on both sides to get:
𝑐3 = 0.01, 𝑐4 = 0, 𝑐5 = −0.002

General solution is then: 𝑦(𝑥) = 𝑐1 cos(𝑥) + 𝑐2 sin(𝑥) + 0.01𝑥 2 − 0.002

Now apply the BCs to find c1 and c2:
𝑦(𝑥) = 0.002 cos(𝑥) + 1.45 sin(𝑥) + 0.01𝑥 2 − 0.002

b. 𝑦 ′′ + 3𝑦 ′ + 2.25𝑦 = −10𝑒 −1.5𝑥 BCs: 𝑦(0) = 1, 𝑦 ′ (5) = 0

This is also a second order non-homogeneous constant coefficient ODE. Again, sum the
complementary and particular solutions.

𝑦𝑐 = (𝑐1 + 𝑐2 𝑥)𝑒 −1.5𝑥

𝑦𝑝 = 𝑐3 𝑥 2 𝑒 −1.5𝑥

First substitute yp into the ODE to find c3 = –5

General solution is then:

𝑦(𝑥) = (𝑐1 + 𝑐2 𝑥 − 5𝑥 2 )𝑒 −1.5𝑥

1
And apply to BCs to find c1 and c2:
𝑦(𝑥) = (1 + 20.92𝑥 − 5𝑥 2 )𝑒 −1.5𝑥

PROBLEM 2

The governing equation for the steady state temperature profile within the pipe (as a function of
radius) is:

1𝑑 𝑑𝑇
(𝑟 ) = 0
𝑟 𝑑𝑟 𝑑𝑟

The heat transfer zone is between Ri and Ro, and the boundary conditions at Ri and Ro are:

𝑇(𝑅𝑖 ) = 𝑇𝑖
BCs{ 𝑑𝑇
−𝑘 𝑑𝑟 (𝑅𝑜 ) = ℎ(𝑇(𝑅𝑜 ) − 𝑇∞ )

The reason for these BCs is that inside the pipe, we have condensation, which happens at constant
temperature. So there is no heat transfer within the fluid inside the pipe, and the temperature at the
inner wall will therefore be equal to the temperature of the steam. At the outer surface, we have the
matched flux condition where heat brought to the surface by conduction is equal to the heat removed
by convection.

Integrating the field equation twice gives the general solution:

𝑇(𝑟) = 𝑐1 ln(𝑟) + 𝑐2

Apply the BCs to find the two constants c1 and c2 (I STRONGLY encourage you to do this so that
you fully familiarize yourself with the matched flux conditions and can solve such problems on an
exam).

ℎ(𝑇∞ − 𝑇𝑖 ) 𝑟
𝑇(𝑟) = ln ( ) + 𝑇𝑖
ℎln(𝑅𝑜 ⁄𝑅𝑖 ) + 𝑘⁄𝑅𝑜 𝑅𝑖

2
Bonus discussion: plotting the temperature profile is quite informative. Given the constants in the
problem, a plot of temperature from Ri to Ro looks like:

Observations:
1) The temperature barely changes between the inner and outer surfaces of the pipe (it goes
from 573 K to 572.88 K). Question: is there a way you could have predicted this feature
(that the temperature will barely vary across the wall thickness) beforehand? Answer:
absolutely, by calculating the Biot number. This is basically a situation that would be a good
candidate for a “lumped heat capacitance” model.
2) The temperature profile looks almost linear, even though the equation was completely
nonlinear in r (it involved a natural logarithm dependence). Question: can you guess why
this is the case? Answer: this is because the wall is relatively thin (compared to its radius)
so curvature can almost be ignored (just like you don’t see trees and buildings as being
diverging as they get taller and taller, even though they are growing on a curved surface.
The height of a building, being analogous to the thickness of the wall in this problem, is
much much smaller than the radius of curvature that it is sitting on).

Just for fun, here is the comparison of temperature profiles between two cases: a) k = 400 W/mK
and h = 5 W/m2K as given in the problem statement (shown in red), and b) k = 200 W/mK and
h = 1000 W/m2K (shown in blue). As you can see, the second case will result in a huge temperature
drop of close to 200 K across the tube wall.

3
b. The rate at which heat is getting transferred from the pipe to the outside environment can be
calculated in three different ways. I will present one of them, and strongly encourage you to think
about the other two:

𝑄̇ = 𝑞𝑐𝑜𝑛𝑣 𝐴 = ℎ(𝑇𝑜 − 𝑇∞ )𝜋𝐷𝐿

where To is the temperature at the outer wall, evaluated from the temperature profile.

Plug all the numbers in to get:

20𝑊
𝑄̇ = ( 2 ) (572.9𝐾 − 293.15𝐾)(𝜋)(0.32𝑚)(60𝑚)
𝑚 𝐾
𝐽
𝑄̇ = 337,311⁡ = 337.31⁡𝑘𝑊
𝑠
c. Temperature at the outer surface of the pipe. From the previous parts,

ℎ(𝑇∞ − 𝑇𝑖 ) 𝑅𝑜
𝑇𝑜 = ⁡𝑇(𝑅𝑜 ) = ln ( ) + 𝑇𝑖
ℎln(𝑅𝑜 ⁄𝑅𝑖 ) + 𝑘⁄𝑅𝑜 𝑅𝑖

𝑇𝑜 = ⁡572.88⁡𝐾

d. This is a steady state process. So the same amount of heat that is getting transferred to the outside
environment is getting removed from the saturated vapor inside the pipe, condensing it into saturated
liquid. The condensation happens at 300℃, and the amount of heat removed per unit mass of the
steam is the heat of vaporization (or change in enthalpy between saturated vapor and saturated
liquid) at 300℃.
𝑘𝐽
̂𝑣𝑎𝑝 = 1404.9⁡
From steam tables: ∆𝐻
𝑘𝑔

𝑄̇ 𝑘𝑔
Therefore the rate of condensation is: 𝑚̇ = ̂𝑣𝑎𝑝
= 0.24⁡
∆𝐻 𝑠

4
PROBLEM 3
a. Yes, steady state is a good assumption. This is because the goal of this design is to just eliminate
any heat loss from the inside environment, which means its temperature will not change with time.
Therefore, we are looking at a steady state scenario where there is no heat loss from the inside
environment.
b. Hopefully you recognized that we cannot apply the “sum of resistances” concept here, because of
the heat generation term within the heater.
c. Let us start with the inside environment, as suggested. Note that we are trying to keep heat from
leaving the inside environment under steady state conditions. If heat were to leave this environment,
it would leave with a convective flux given by:

qx,inside = hi (Ti - Ts )
where Ts is the wall temperature at the interface between the wall and the inside environment. Since
we want to have qx,inside = 0 (eliminate any heat loss from inside), we must have: Ts = Ti
This means that the temperature profile in the inside environment must be completely flat at Ti.
This makes sense. If it weren’t flat, it would create a driving force for heat transfer.
d. Now let us focus on the wall itself, as suggested. The governing differential equation for the
wall layer is:

d 2T
=0
dx 2
Therefore the temperature profile is linear in the wall. We know that at the interface with the inside
fluid, we have T = Ti, as just established above. Also we cannot have a steady heat flux through the
wall, because this would lead to temperature changes in the inside environment. Therefore, the BCs
to solve for the temperature profile in the wall are:

ì T (- L ) = T
ïï w i

í dT
ï -k =0
ïî dz 0

These two conditions tell us that the temperature is flat at Ti inside the wall as well.

Twall ( x ) = Ti -Lw < x < 0

e. Now we can move on to the heater layer. In this layer, the governing equation can be derived by
simplifying the general heat transfer equation or writing a shell balance. Either way, you should
get:

d 2T
0 = k H 2 + HV
dx

5
This has the general solution:

HV x 2
T ( x) = - + c1x + c2
2k H

f. At any interface in heat transfer, we must have continuity in temperature and in heat fluxes.
Therefore, at x = 0, we must have:

( ) ( )
T 0- = T 0+ ( )
and qx 0- = qx 0+ ( )
Here, the notations 0– and 0+ mean “on the negative side of x = 0”, and “on the positive side of x =
0”, respectively.
Therefore, having already solved for the wall layer, the conditions written above translate to:

( )
T 0+ = Ti and - k
dT
dx
= -k
dT
dx
=0
0+ 0-
g. Again, we must have temperature and flux continuity. Therefore:

( ) ( )
T L-H = T L+H ( )
and qx L-H = qx L+H ( )
Given that we have a convective environment outside, the conditions written above translate to:

( ) ( )
T L-H = T L+H and - k
dT
dx
( )
= ho éëT L+H - To ùû
L-H

To summarize and write these conditions in clean form, we have:

ì dT
ï T ( 0 ) = Ti and ( 0) = 0
ï dx
BCs í
ï -k dT ( L ) = h éT ( L ) - T ù
ïî dx H ë H oû

Note that we have two conditions written at x = 0 and one at x = LH, for a total of three conditions.
This may initially seem odd to you, because we have a second order ODE so only two BCs must be
needed. But here we not only want to find the temperature profile, we have an additional constraint
on the system, and that is to make sure there is no heat loss from the wall all the way to the inside
environment.

h. We should choose one of the two conditions written at x = 0 (whichever we like) and the one
condition written at x = LH to find the temperature profile. Then, we apply the other condition written
at x = 0 which we haven’t used, to relate Hv to other parameters in the next part.

6
I chose the following BCs to solve for the temperature profile:

ì dT
ï
ï dx
( 0) = 0
BCs í
ï -k dT ( L ) = h éT ( L ) - T ù
ïî dx H ë H oû

Which gave me:

HV x 2
TH ( x ) = - + Ti
2k H

()
i. Now we can apply the other condition at x = 0 (which is T 0 = Ti ) to finish the problem. Once
you do this, you will find the heat generation term must be:

2k H ho (Ti - To )
HV =
LH ( 2k H + LH ho )

PROBLEM 4

In all the calculations below, we are assuming the mixture is an Ideal Gas, which is justified by the
low total pressure.

a. 𝑝𝐶𝑂2 = 𝑦𝐶𝑂2 𝑃 = (0.9532)(6.10 × 10–3 ⁡𝑏𝑎𝑟) = 5.81 × 10–3 ⁡𝑏𝑎𝑟 = 518⁡𝑃𝑎

𝑝𝐶𝑂2 𝑚𝑜𝑙
b. 𝐶𝐶𝑂2 = = 0.333⁡
𝑅𝑇 𝑚3

𝑃 𝑚𝑜𝑙
c. 𝐶 = 𝑅𝑇 = 0.349⁡ 𝑚3

d. Taking 1 m3 of the mixture as basis, we can calculate the mass:

𝑁

𝑚 = ∑ ℕ𝑖 𝑀𝑖
𝑖=1

Where ℕ𝑖 and 𝑀𝑖 denote the number of moles and the molecular weight of species i. Since we have
1 m3 of the mixture, the total number of moles in, ℕ, is known, and equal to 0.349 mol (from part c)

𝑚 = ⁡ 𝑦1 ℕ𝑀1 + ⁡ 𝑦2 ℕ𝑀2 + 𝑦3 ℕ𝑀3 + 𝑦4 ℕ𝑀4 + 𝑦5 ℕ𝑀5 = 15.2⁡𝑔

Since we had 1 m3 of the mixture, the density is then:

𝑔
𝜌 = 15.2⁡
𝑚3

7
PROBLEM 5

a. This scenario involves the movement of charged species because they are getting
“pushed” by an external electric field (and the species are charged). The movement of
charged species in such a case, termed Migration in electrochemical engineering, is not
due to entropy and the minimization of Gibbs free energy or a gradient in chemical
potential as we discussed in class. Migration is due to an externally applied electrical
potential and the electromagnetic force that it exerts on charged species.
b. The movement of gas molecules in this case is due to a pressure drop, and not a gradient
in their chemical potential. This is the topic of CBE120A Momentum Transfer.
c. Yes, there is a gradient in the concentration of oxygen, nitrogen, and all other trace
molecules: they are more concentrated near the earth’s surface. But this is because there is
an external gravitational potential exerting gravitational force on them. The conditions that
arise (more molecules near the earth’s surface than far away from the earth) is already an
equilibrium condition, resulting from the balance of entropy and gravitational potential.
Therefore, the system will remain at these equilibrium conditions, and the molecules will
not start to disperse and move away from near the earth’s surface. Again, the reason is that
there is an external potential acting on the molecules (similar to Case a above), which,
over the very large distances considered (from here to the outer layers of the stratosphere,
for example), results in a gradient in gas concentrations under equilibrium conditions.