Profa.

Liliane Lona – Em revisão

Chapter 4 – Distributed-Parameters Models

Different from the previous chapter, that hasstudied lumped-parametersproblems, this
chapter deals with examples in which variables such as concentration and temperature will
vary alongthe position, characteristic of distributed-parameters problems.

Section 4.1 brings some simple introductory examples needed to understand how to model
distributed parameter problems. Section 4.2 presents some concepts about transport by
diffusionand models more complex systems. Finally, section 4.3 presents some examples with
variation in more than one spatial dimension.

4.1 – Some introductory examples

Example 4.1: Assume a fluid at 200C fed in a cylindrical tube of length (L) 60m and radius
(R)0.2m in a rate (Q) 4m3/h.Assume also this tube exchanges heat with a jacket, whose
temperature is 3000C. Determine the axial profile of temperature inside the tube. Consider the
system is in steady state and that there is no radial or angular variation of temperature inside
the tube. Consider also thatthe thermal diffusion is not important in any direction (axial, radial
and angular).

Solution:

According to the recipe presented in chapter 2, the first thing to do in order to model this tube
is to define the controlvolume. Differently from examples presented in chapters 3, the
temperature of the fluid is not the same along the tube, so the entire tube cannot be
considered as control volume. In this way, we will consider a very small slice of the tube, of
length ∆x, and we will assume that in this small slice the fluid temperature is the same (see
Figure 4.1).

Figure 4.1: Fluid flowing in a tube in a steady state regime.

The conservation law (E – L + G – C = A) will be applied to this small slice (control volume) in
order to obtain the energy balance. This energy balance will be valid for any slice taken in any
position inside the tube, so we can say that the energy balance obtained for the slice of tube
shown in Figure 4.1, will be valid for the entire tube.

1
 Profa. Liliane Lona – Em revisão

In our problem, there is neither generation nor consumption of energy and the system is in
steady state, so the conservation law equation becomes:

E–L=0 (4.1)

Let us now define the energy that enters and leaves the control volume. There is heat
exchange by convection with the jacket, and as defined in chapter 3, this flow of energy can be
expressed by:

Convection heat transfer rate = U x A x (Tw - T), in which

U = global coefficient of heat transfer [J / h m20C]

A = area of heat exchange in the control volume (𝐴𝐴 = 2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 [m2])

T = temperature inside the control volume [0C]

Tw = temperature of the jacket [0C]

So,

The Flow of Energy by Convection = U x 2 𝜋𝜋𝜋𝜋 ∆𝑥𝑥 x (Tw - T) [J/h] (4.2)

Observe that, as define in chapter 3, we will add this term in the energy balance, because this
term will be either negative (heat leaving the system) if the environment temperature is lower
than the temperature of the system, or positive (heat being added to the system) if the
environment temperature is higher than the temperature of the system. In this way, we do not
need to worry defining if this energy enters (E) or leaves (L) the controlvolume.

Figure 4.1 shows that the fluid enters the control volume in the position x and leaves in the
point x + ∆x.

The amount of energy entering the control volume at the point x is given by:

𝑚𝑚 3 𝑘𝑘𝑘𝑘 𝐽𝐽
𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑡𝑡ℎ𝑎𝑎𝑎𝑎 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑖𝑖𝑖𝑖 𝑥𝑥: 𝑄𝑄 × 𝜌𝜌 × 𝐶𝐶𝐶𝐶 × 𝑇𝑇 0𝐶𝐶 ,
ℎ 𝑚𝑚 3 𝑘𝑘𝑘𝑘 0 𝐶𝐶

so:

𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎: 𝑄𝑄 𝜌𝜌𝜌𝜌𝜌𝜌𝑇𝑇 [J/h] (4.3)

Be careful: Observe that we do not know the temperature of the fluidentering in the
positionx,therefore we write a generic temperature T. 200C is the temperature of the fluid at
position 0, and not at position x.

In order to obtain the amount of energy that leaves the control volume in the point x + ∆x we
use the concept of the infinitesimal variation of the dependent variable withthe independent
variable, as shown below:

2
 Profa. Liliane Lona – Em revisão

x (enters) x + ∆x (leaves) Dimension

𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇) J/h
𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑥𝑥
𝑑𝑑𝑑𝑑

𝑑𝑑(𝑄𝑄 𝜌𝜌𝜌𝜌𝜌𝜌𝜌𝜌 )
𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻ℎ𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎𝑎 𝑖𝑖𝑖𝑖 𝑥𝑥 + ∆𝑥𝑥: 𝑄𝑄 𝜌𝜌𝜌𝜌𝜌𝜌𝜌𝜌 + 𝑑𝑑𝑑𝑑
∆𝑥𝑥 [J/h] (4.4)

Using the expression (4.2), (4.3) and (4.4) in the equation of conservation law (4.1) yields:

𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇)

𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 − ( 𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑥𝑥) +U2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 (Tw - T) = 0
𝑑𝑑𝑑𝑑

or

𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇)

− + 2 𝜋𝜋 𝑅𝑅 U �Tw – T� = 0 (4.5)
𝑑𝑑𝑑𝑑

Observe that, as the system is in the steady state, it was not necessary to consider the amount
of energy that enters and leaves the control volume in a period ∆t. This is because there is no
accumulation of energy over time (the temperature profile along the tube length does not
change with the time).

If we assume that the 𝜌𝜌 and Cp do not vary with the temperature (and consequently with the
length), the Equation (4.5) becomes (remember Tw = 3000C):

𝑑𝑑 𝑇𝑇 2 𝜋𝜋 𝑅𝑅 U
= (300 – T)
𝑑𝑑𝑑𝑑 𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶

With the initial condition: at x = 0, T = 200C.

In order to analyze the temperature profile obtained for this system, let us assume:density (𝜌𝜌)
= 900kg/m3, specific heat of the fluid (Cp) = 3000 J/kg0C, and the global coefficient of heat
transfer (U)= 60000 J / h m20C.The profile of temperature along the length can be seen in the
Figure 4.2 (Chapter 6 will show how to solve this ODE using Excel).

Analyzing the temperature profile,it can be observed that the temperature of the fluid is 200C
when fed in the tube, and starts increasing along the length due to the heatexchange with the
jacket that is at 3000C. Observe that at the end of the tube (60m) the temperature of the fluid
reaches 1160C. If the tube were long enough, the temperature of the fluid would tend to
3000C, but it is impossible to reach temperatures higher than 3000C inside the tube.

3
 Profa. Liliane Lona – Em revisão

Figure 4.2: Profile of the temperature inside the tube.

Despite the fluid temperature changes along the length, the system is in steady state, because
the axial profile of temperature does not change over time.

Example 4.2: Let us imagine now that, for some reason, the temperature of the jacket in the
example 4.1 changes instantaneously and abruptly from 3000C to 2000C. The system,which was
in a steady state,suffers a modification, and the profile of temperature inside the tube will
change with time(the temperature will decrease) until the system reaches a new steady state.
How would be the profile of temperature of the fluid inside the tube along the length and over
time until the new steady stateis reached? Solve this problem assuming neither radial no
angular temperature profiles, and no heat diffusion in any direction inside the tube.

Solution: If the temperature of the jacket changes abruptly from 3000C to 2000C, the profile of
temperature shown in Figure 4.2 will be modified. The curve will start in 200C, because the
fluid is fed in the tube at this temperature, but the fluid temperature will increase less,
because the jacket is 1000C colder. One can imagine that the axial profile of the temperature
inside the tube will start as shown in Figure 4.2 (at time = 0), but it will change over time until a
new steady state is reached.

The modeling of this system will generate a Partial Differential Equation (PDE), because the
temperature will change along the tube and over the time.

The same slice of the tube will be considered as control volume, because this is a distributed
parameter problem. As there is temperature variation with time, the accumulation term must
be considered, and the conservation law equation becomes:

E–L=A

Also in this case, there is no generation or consumption of energy, because there is no

chemical reaction. Comparing to Equation 4.1 one can observe that the accumulation term
was added.

4
 Profa. Liliane Lona – Em revisão

The expressions to represent the amount of energy that enters atx and leaves atx + ∆x, as well
as the convectionterm are calculated as per example 4.1 and are rewrite below:

𝐻𝐻𝑒𝑒𝑒𝑒𝑒𝑒 𝑡𝑡ℎ𝑎𝑎𝑎𝑎 𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒𝑒 𝑖𝑖𝑖𝑖 𝑥𝑥: 𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 [J/s] (4.6)

𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇)

𝐻𝐻𝐻𝐻𝐻𝐻𝐻𝐻 𝑡𝑡ℎ𝑎𝑎𝑎𝑎 𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 𝑖𝑖𝑖𝑖 𝑥𝑥 + ∆𝑥𝑥: 𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑥𝑥 [J/s] (4.7)
𝑑𝑑𝑑𝑑

The Flow of Convection Energy = U 2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 (Tw - T) [J/s] (4.8)

Besides these three terms, we need to calculate the amount of energy accumulated in the
control volume. We will do that asit was donein the previous chapter, i.e. by calculating the
amount of energy accumulated in a very short period of time ∆t, which is the energy in the
time 𝑡𝑡 + ∆𝑡𝑡 minus the energy in a time t, see below.

t t + ∆t Dimension
𝑉𝑉 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 𝑑𝑑(𝑉𝑉 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇) J
𝑉𝑉 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑡𝑡
𝑑𝑑𝑑𝑑

Once again, we assume that 𝜌𝜌 and Cp do not change with temperature (remember that the
temperature is changing with time, and if 𝜌𝜌 and Cp do not change with temperature, they will
not vary over time), so the accumulation term becomes:
𝑑𝑑 𝑇𝑇
Accumulation term = 𝑉𝑉 𝜌𝜌 𝐶𝐶𝐶𝐶 ∆𝑡𝑡 (J)
𝑑𝑑𝑑𝑑

The volume in the accumulation term is the control volume (V = 𝜋𝜋R2∆x).

Since the accumulation term represents the amount of energy that is accumulated in a period
of time ∆t, we must also consider the energy that enters and leaves the control volume in this
same period of time, so we must multiply the expressions (4.6), (4.7) and (4.8) by ∆t.

The energy balance, in a period∆t becomes:

𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇) 𝑑𝑑 𝑇𝑇

𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇∆t - (𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑥𝑥)∆𝑡𝑡 +U 2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 (Tw - T) ∆t = 𝑉𝑉 𝜌𝜌 𝐶𝐶𝐶𝐶 ∆𝑡𝑡
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑

or
𝑑𝑑𝑑𝑑 𝑑𝑑 𝑇𝑇
−𝑄𝑄𝑄𝑄 𝐶𝐶𝐶𝐶 ∆𝑥𝑥∆𝑡𝑡 + U 2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 (Tw - T)∆t = 𝜋𝜋R2∆x 𝜌𝜌 𝐶𝐶𝐶𝐶 ∆𝑡𝑡
𝑑𝑑𝑑𝑑 𝑑𝑑𝑑𝑑

Observe that ∆t and ∆𝑥𝑥 appear in all terms of the energy balance and can be simplified. In fact,
this simplification always must happen when developing a model. If at this point we were not
able of cutting all ∆´s of our balance, it is because we made some mistake, and our model has
to be double checked. Observe also that there are two independent variables (x and t), so the
symbol dmust be changed to 𝜕𝜕. So, the energy balance becomes:

𝜕𝜕𝜕𝜕 2 𝑈𝑈 𝑄𝑄 𝜕𝜕𝜕𝜕
= (𝑇𝑇𝑇𝑇 − 𝑇𝑇) − (4.9)
𝜕𝜕𝜕𝜕 𝑅𝑅𝑅𝑅𝑅𝑅𝑅𝑅 𝜋𝜋𝑅𝑅 2 𝜕𝜕𝜕𝜕

This PDE needs two boundary or initial conditions, one related to the length and other to the
time.

5
 Profa. Liliane Lona – Em revisão

The initial condition related to the time can be obtained remembering that, at the beginning
(at t = 0h), the profile of temperature inside the tube is the one shown in Figure 4.2, because
the system operates like that before the change in the jacket temperature. The curve in the
Figure 4.2 can be represented by a function that correlates the temperature and the length
(T=f(x)).

The initial condition related to the length is obtained remembering that the temperature of
the fluid fed in the tube is 200C, so, at x = 0, T = 200C.

After solving Equation 4.9 using these two initial conditions(details of numerical solution will
be presented in Chapter 7), the following temperature profiles along thelength and over
thetime can be obtained (see Figure 4.3):

Figure 4.3: Profiles of temperature of the fluid along the length and over the time.

One can observe that the axial profile of temperature varies over time, with greater variations
in the beginning (see the difference of the profiles from 0 to 0.8h), and minor as the time goes
on (see the small difference in the curves from time 1.6h to 2.4h), until a new steady state is
reached, when the axial profile of temperature does not change anymore with time. In our
case, after 2.4h the temperature profile along the length remains the same, so we can say that
the new steady state was achieved after 2.4h.

Example 4.3 – This problem is adapted from the book Process Heat Transfer, by Q. D.
Kern(1950), and develops amodel to designbitubular heat transfer. Let us consider now two
concentric tubes, as shown in Figure 4.4, with benzene flowing through the internal tube and
toluene flowing through the annulus. The two fluids flow in parallel and the system is in steady
state. The benzene and toluene are fed in a rate 9820lh/h (Wbez) and 6330lb/h (Wtol)
respectively. These concentric tubes are used to increase the temperature of benzene from 60
to 1000F and reduce the temperature of toluene from 170 until 1100F.Assume that the toluene
does not exchange heat with the environment, just with the benzene, so that the heat
exchange occurs through the wall of the internal tube that has 1.25in of diameter (ignore the
thickness of the tube). What is the required length of the tubes to make the thermal exchange
needed? Let us assume that Cp of benzene (Cpben) and toluene (Cptol) are 0.425 Btu/(lb) (0F)

6
 Profa. Liliane Lona – Em revisão

and 0.440 Btu/(lb) (0F) respectively and that these values do not change significantly with the
temperature.

Figure 4.4: Double-pipe heat exchanger operating with parallel flow

Solution:The solution of this problem is like the solution of problem 4.1, but in this case, it is
necessary to make the energy balance for both fluids, i.e., for the benzene flowing in the tube
and for the toluene flowing in the annulus. We will first present a solution similar to what we
have done so far, but in the sequence, some assumptions will be made, in order to bring the
solution of this problem closer to what is done in the design of heat exchange equipment.

Heat exchangers are very common in chemical industries and they are very useful to exchange
heat between different streams in a plant. This equipment can assume many different
geometries and the one considered in this example is the simplest (bitubular heat exchanger
with parallel flow), although the calculus presented herein is not very different from the
calculus for more complex heat exchangers.

To model this system, we must define a control volume. As this is a distributed parameter
problem (temperature changes along length), an increment ∆x is considered and the energy
balance for both fluids is done considering this small volume.

x x + ∆x Dimension
Benzene WbenCpbenTben 𝑑𝑑(𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 ) Btu / h
𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 + ∆𝑥𝑥
𝑑𝑑𝑑𝑑

Toluene WtolCptolTtol 𝑑𝑑(𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 ) Btu / h

𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 + ∆𝑥𝑥
𝑑𝑑𝑑𝑑

The amount of heat exchanged by convection is written the same way as done before, keeping
in mind the sign convention adopted.

Heat exchanged by convection (Btu / h)

Benzene U A (Ttol – Tben)

Toluene U A (Tben – Ttol)

In which:

U = global coefficient of heat transfer (Assume U = 0.8 [Btu / h in20F])

7
 Profa. Liliane Lona – Em revisão

A = area of heat exchange in the volume of control (𝐴𝐴 = 2 𝜋𝜋 𝑅𝑅 ∆𝑥𝑥 ⇒ 𝐴𝐴 = 1.25 𝜋𝜋 ∆𝑥𝑥 ) [in2]

The conservation law for benzene and toluene yields: E – S = 0, and the energy balance for
both fluids can be written as:

Benzene:

𝑑𝑑(𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 )

𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 − (𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 + ∆𝑥𝑥) + 𝑈𝑈 1.25 𝜋𝜋 ∆𝑥𝑥(𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 − 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 )
𝑑𝑑𝑑𝑑
=0

Toluene:

𝑑𝑑(𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 )

𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 − (𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 + ∆𝑥𝑥) + 𝑈𝑈 1.25 𝜋𝜋 ∆𝑥𝑥(𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 − 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 ) = 0
𝑑𝑑𝑑𝑑

After simplifying terms, we obtain:

𝑑𝑑𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏
Benzene: 𝑊𝑊𝑏𝑏𝑏𝑏𝑏𝑏 𝐶𝐶𝐶𝐶𝑏𝑏𝑏𝑏𝑏𝑏
𝑑𝑑𝑑𝑑
= 𝑈𝑈 1.25 𝜋𝜋(𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 − 𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 ),initial condition: at x = 0, Tben = 600F

𝑑𝑑𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡
Toluene: 𝑊𝑊𝑡𝑡𝑡𝑡𝑡𝑡 𝐶𝐶𝐶𝐶𝑡𝑡𝑡𝑡𝑡𝑡 = 𝑈𝑈 1.25 𝜋𝜋 (𝑇𝑇𝑏𝑏𝑏𝑏𝑏𝑏 − 𝑇𝑇𝑡𝑡𝑡𝑡𝑡𝑡 ), initial condition: at x = 0, Ttol = 1700F
𝑑𝑑𝑑𝑑

Considering numerical values for all parameters (Wbez = 9820lb/h, Wtol= 6330lb/h, Cpben = 0.425
Btu/(lb) (0F), Cptol = 0.440 Btu/(lb) (0F) and U = 0.8 [Btu / h in20F]), this system of two Ordinary
Differential Equations can be solved numerically (details of numerical solution will be
presented in Chapter 6), to generate the temperature profiles for benzene and toluene along
the length of the tubes, as shown in Figure 4.5.

Figure 4.5: Temperature profiles inside the bitubular heat exchanger.

One can observe that the benzene and toluene reach the desired temperature (Tben= 100 e
Ttol= 1100F) when the length of the tubes is around 1290 inches (or 110 feet).

To design a heat exchanger, usually many simplifying hypotheses are made,to make the
numerical solution easier and faster. Typically, the design of this equipment means finding the
size (area) of the heat exchanger able to make the desired changes in the temperature

8
 Profa. Liliane Lona – Em revisão

between hot and cold fluids. Assuming there is no heat loss to the environment, the amountof
heat received by the cold fluid is equal to the one given by the hot fluid, and this amount
isequal to the energy exchanged between the two fluids by convection. The amount of energy
exchanged by convection varies along the length of the tube, because the difference of
temperature between the hot and cold fluid (driving force) varies along the length (see Figure
4.5).

In the design of heat transfer equipment,the Logarithmic Mean Temperature Difference

(LMTD) is considered to simplify numerical solution. The LMTD is an average of the difference
of temperature between the two fluids along the entire equipment (average driving force
along the entire equipment), and is given by:

In which ΔTA is the difference of temperature between the two streams at the terminal
wherethe hot fluid is fed (in this case: 170 - 60 = 1100F) and the ΔTB is the difference of
temperature between the two streams at the other terminal (in this case: 110 – 100 = 100F). In
our case, LMTD is 41.70F.

The total amount of heat received by the cold fluid and given by the hot fluid is WbenCpben(100-
60) = WtolCptol(170-110) = 1.67x105, and this amount is equal to the heat exchanged by
convection between the two fluids (U x A xLMTD).

So: A = 1.67x105/ (U x MLDT)⇒ A = 1.67x105/ (0.8 x 41.7)⇒ A = 5006 in2

The heat exchange area of the tube (A) can be written as 𝜋𝜋𝜋𝜋𝜋𝜋, in which D is the diameter of
the tube (1.25in) and L is its length. In this way, the length of the tube can be find as: L = A/𝜋𝜋𝜋𝜋,
or L = 1275in.

Observe that the value obtained using this simplified calculation (L = 1275in) is only 1.2%
lowerthan the one obtained when the system of two ODEs are solved together (L = 1290in).
For this reason, the project of heat exchanger equipment is usually done as per the second
approach; however, the recipe presented in this book can also be used to design heat
exchangers.

After these three simple introductory examples on distributed-parameter problems, section

4.2 will revisit some concepts on mass, energy and momentum transfer by diffusion, needed to
model more complex systems. The idea is just to show a few pieces of information needed to
model systems in which transfers by diffusion are relevant. The reader can find more details on
this subject in specific literature dealing with transfer phenomena (Bird et al. (2007), Welty et
al. (2007) and Bergman et al. (2011), just to mention a few).

4.2 – Concepts about transfer by diffusion

All flow rates (mass, heat and momentum) is directly proportional to a driving force and
inversely proportional to a resistance:

Flow Rate α Driving Force

9
 Profa. Liliane Lona – Em revisão

Resistance

We used this concept in chapter 3 (section 3.2) for heat flow, but it is also valid for mass and
momentum. This section will revisit some concepts about diffusive transportand will point out
some analogies among transfers of heat, mass and momentum.

4.2.1 - Diffusive Transport of Heat:

Imagine a solid cube of 1m3 initially at 500C(Figure 4.6a). Two opposite sides of the cube are
submitted to 70 and 300C (assume that both temperatures do not vary over time) and all other
faces are isolated, so there is energy flow only in the x direction.

One can imagine that as the energy starts flowing from the hotter to the colder face, the
internal profile of temperature inside the cube starts changing until the system reaches the
steady state.Figure 4.6b shows an example of how the axial profiles of temperature inside the
cube could change over time. Since there is no heat exchange with the environment, at the
steady state the temperature profile inside the cube is given by a straight line.

Figure 4.6a: Solid cube, initially at 500C, Figure 4.6b: Profiles of temperature along the
with 4 isolated faces and two opposite length of the cube over time
faces at 70 and 300C

The energy flow inside the cube is only due to heat diffusion (heat conduction from molecule
to molecule). Fourier observed that the flow of energy by diffusion can be expressed by:

q dT
( )x = −k ( ) (4.10)
A dx

In which (considering international units):

q = energy flow (J/s)

A = cross section from where the energy flows (m2)

(q/A)x = energy flux in the x-direction (J/sm2)

k = thermal conductivity (J/s m 0C)

10
 Profa. Liliane Lona – Em revisão

dT/dx = temperature gradient (0C/m)

Observe that the energy flux (q/A)x of the Fourier equation is directly proportional to driving
force (dT) and inversely proportional to a resistance (dx/k).

The negative sign of Equation (4.10) is due to the orientation of q, dx and dT: Figure 4.6b
shows that dx is positive and dT is negative, making dT/dx negative. Since the energy flux
(q/A)x is positive (grows in the same sense as x), the minus sign is necessary to make the
equation (4.10) coherent.

4.2.2 - Diffusive Transport of Mass:

Imagine a cubic box of 1m3 open at the top, with only air inside it. At some point, gas ethylene
at a constant concentration C = 1 mol/m3 starts blowing through the box, as shown in Figure
4.7a. Imagine that, for some mechanism chemically possible, the concentration of ethylene at
the bottom of the box is always zero.

Figure 4.7a: Cubic box initially with air. Figure 4.7b: Profiles of ethylene concentration
Ethylene concentration at the bottom along the cube height over time
and at the top equal to 0 and 1mol/m3
respectively.

In the begging, the concentration of ethylene inside the box is zero, but, as the ethylene starts
blowing perpendicularly through the box, there will be a diffusive mass flow of ethylene from
the region with higher to the region with lower concentration of ethylene (the driving force is
the difference of concentration). This diffusive mass flow will generate axial profiles of ethylene
concentration inside the box,which will vary over time until the steady state is reached(see
Figure 4.7b). In this example, we assume that there is no profile of concentrating along y and z.

In 1855, Fick experimentally obtained theEquation (4.11) to represent the molar flow of some
compound (say A) by diffusion.

JA dC
( ) x = − D( A )
A dx (4.11)

11
 Profa. Liliane Lona – Em revisão

In which (considering international units):

JA = molar flow of component A (mol/s)

A = cross-sectionalarea from where the molecules of component A flows (m2)

(JA/A)x = molar flux of component A in the mass flow direction (mol/m2s)

D = diffusion coefficient or diffusivity of component A (m2/s)

dCA/dx = concentration gradient for the component A (mol/m3m)

In this case it is also observed that the molar flux of component A (JA/A)x of the Fick´s law is
directly proportional to driving force (dCA) and inversely proportional to a resistance (dx/D).

Analogous to the Equation (4.10), the negative sign of Fick´s law is necessary to make the
Equation (4.11) coherent with the orientation of JA, dx and dCA.

4.2.3 - Diffusive Transport of Momentum:

For didactical point of view, the example presented is this section considers a Newtonian Fluid.
Fluids can be classified as Newtonian and Non-Newtonian, but definitions and examples about
the different types of fluids are not the scope of this book, and can be found in literature
related to Mechanic of Fluids and Transport Phenomena (Bird et al. (2007), Fox et al. (2012)
and Welty et al. (2007)).

In our example, imagine a thin layer of a fluid between two long plates, as shown in Figure
4.8a. The thickness of the fluid layer is d(m). Initially (t=0) the system is at rest, and after that,
the upper plate starts moving to the positive direction of x at velocity U equal toUx (m/s).
Consider that the lower plate keeps set (U= 0).

In the beginning (t=0) the fluid velocity is zero. When the movement starts, the molecules of
the fluid in contact with the plates assume the same velocity of the plates, and due to the
friction forces, a velocity profile is formed, as shown in Figure 4.8b. As the time goes on, more
and more fluid is drawn toward the moving plate.

Newton observed that, in laminar flow, when the steady state is reached, there is a linear
velocity profile inside the fluid layer (Newton´s law of viscosity), as can be observed in Figure
4.8b.

12
 Profa. Liliane Lona – Em revisão

Figure 4.8a: Thin layer of a Newtonian Figure 4.8b: Profiles of velocity along the
fluid between two long plates layer height over time
If the force applied to the upper plate is “F” and the area of the upper plate is “A”, the ratio
“F/A” is known as shear stress, and it is equal in magnitude to the momentumflux.Physically,
the momentum flux is the transfer of momentum through the fluid from a region with higher
to other with lower velocity, and can be represented by τ yx , in which y is the direction of the
transfer of momentum, and x is the direction of the movement velocity.

In the Newton´s law of viscosity, the momentum flux ( τ yx ) is directly proportional to a driving
force (dU) and inversely proportional to a resistance dy/µ, as can be seen in Equation (4.12):

F dU
= τ yx = − µ ( )
A dy (4.12)

In which (considering international units):

F = force applied to the upper plate (N)

A = area of the upper plate (m2)

τ yx = shear stress (momentum flux) in which y represents the normal component to the plane
of action of the shear stress and x is parallel to the velocity of the plate (N/m2)

µ = viscosity of the fluid (N s/m2)

dU/dy = gradient of velocity (1/s)

For the same reason mentioned before, the minus sign in Equation (4.12) is needed.

4.2.4–Analogies among all Diffusive Transports

Observe that the same behavior is observed for the diffusive transfers of heat,mass and
momentum. Table 4.1 summarizes the analogies among these three kinds of diffusive
transports presented before.

Table 4.1: Analogies among Diffusive Transports

Heat Mass Momentum

Law Fourier Fick Newton

13
 Profa. Liliane Lona – Em revisão

Equation q dT JA dC F dU
(
A
)x = −k (
dx
) (
A
) x = − D( A )
dx
= τ yx = − µ ( )
A dy
Driving Force dT dCA dU
Resistance dx/k dx/D dy/µ

4.2.5 – Example considering the diffusive effects on the modeling

The examples presented in this section take into account problems in which the diffusive
effects are important and cannot be neglected. For didactical point of view, in this section it
will be consider variations in only one direction. The examples 4.4 and 4.7 are analogous to the
schemes presented in sections 4.2.1 and 4.2.2, so it will bepossible visualize the mathematical
models that generate Figures 4.6b and 4.7b.

Example 4.4: Imagine a cylindrical metal bar with length (L) of 1m and radius (R) equal
to0.03m, initially at 500C (Figure 4.9). At some point, the two edges of the bar are fixed in walls
that are at 70 and 300C (assume that the temperature of the walls do not vary with time).
Imagine that this bar is isolated, so that there is no heat transfer by convection between the
bar and the environment. This system is very similar to the one shown in Figure 4.6a. In this
example, we want to know how long it takes to reach the steady state and how would be the
internal profiles of temperature until then.

Figure 4.9: Heat conduction in an isolatedcylindrical bar

Solution: As this is a distributed-parameter problem (there is an axial profile of temperature), a

slice of the cylinder must be considered as control volume, as depicts Figure 4.9, whose
volume is 𝜋𝜋R2∆x.

The conservation law applied to this control volume yields: E – S = A.

Observe that there is neither generation nor consumption of energy in this case, and that
there is no heat transfer by convection with the environment. The accumulation term must be
considered, because the axial profile of temperature will change with time until the steady
state is reached.

The flow of energy (q) inside the cylinder occurs by conduction (from molecule to molecule of
the metal) due to the difference of temperature (driving force), and can be represented by the
Fourier Law (Equation 4.10), rewritten below:

14
 Profa. Liliane Lona – Em revisão

dT
q = −kA (4.13)
dx

The amount of energy that enter (at x) and leaves (at x+∆x) the control volume by conduction
is shown as follow:

x (enters) (J/s) x + ∆ x (leaves) (J/s)

Energy dT dT d dT
− kπR 2 − kπR 2 + (− kπR 2 )∆x
flowbyconduction dx dx dx dx

In which (considering international units):

k = thermal conductivity (J/s m 0C)

πR2 = cross section area from where the energy flows (m2)

dT/dx = temperature gradient (0C/m)

The accumulation of energy (in Joules) inside the control volume in a period∆t (h) can be
obtained by the difference of energy in a time t+∆t and t.

t t + ∆t
Energy ( J ) ρ × V × Cp × T d ( ρ × V × Cp × T )
ρ × V × Cp × T + ∆t
dt

In which (considering international units):

ρ = density of the metal bar (kg/m3)

V = volume of the control volume = 𝜋𝜋R2∆x (m3)

Cp = specific heat of the metal bar (J/kg 0C)

T = Temperature of the metal bar (0C)

Considering that ρ, Cp and Vof the metal do not change over time, the accumulation of energy
in a period∆t can be written as:

Accumulation of Energy in ∆t (in Joules) = ρπR 2 ∆xCp dT ∆t

dt

Since the amount the energy entering and leaving the control volume is given in J/s and the
accumulation of energy in J, the energy flow must be multiplied by ∆t, to make the units
compatible. Doing that, the energy balance can be written as follow:

dT dT d dT dT
− kπR 2 ∆t − (−kπR 2 + (−kπR 2 )∆x)∆t = ρπR 2 ∆xCp ∆t
dx dx dx dx dt (4.14)

Simplifying terms and considering the thermal conductivity of the metal keeps constant, the
following partial differential equation (PDE) is obtained:

15
 Profa. Liliane Lona – Em revisão

∂T k ∂ 2T
=
dt ρCp ∂x 2 (4.15)

To solve this PDE, three initial or boundary conditions are needed: two related to the space
and one related to the time. Remember that in the beginning (at t = 0), the entire bar is at
500C, and after that, the two edges of the bar are kept at 70 and 300C (left and right). So, the
three conditions are:

At t = 0h, T = 500C, for 0 ≤ L≤ 1m

At x = 0m, T = 700C, for t> 0h

At x = 1m, T = 300C, for t> 0h

Considering numerical values (k = 398.2 J/s m 0C, Cp = 386.3 J / kg 0C and ρ = 8933 kg / m3), the
PDE can be solved numerically (details in Chapter 7) and can generate axial profiles of
temperature over time, as shown in Figure 4.6b. For the metal with properties considered in
this example, after 10 min the internal profile of temperature inside the bar almost does not
change anymore over time.

If one would like to know only the axial profile of temperature in the steady state, the energy
balance presented in Equation (4.15) could be simplified, because the accumulation term
would not be used, so the final energy balance would be:

d 2T
=0
dx 2 (4.16)

With the conditions:

At x = 0m, T = 700C

At x = 1m, T = 300C

Equation (4.16) can be easily integrated to generate:T = - 40x + 70, which is the linear profile of
temperature at the steady state shown in Figure 4.6b.

Example 4.5: Nowimagine that the cylindrical metal bar of the example 4.4 is initially at 500C,
but, this time, only one edge of the bar is fixed in a wall at 700C (constant temperature) and
the other exchange heat with the environmentthat is at 250C (see Figure 4.10). Consider that
the rest of the cylindrical metal bar is isolated (no heat exchange with the environment). Do
again the modeling of this system to obtain how the internal profile of temperature changes
over time.

16
 Profa. Liliane Lona – Em revisão

Tenv=25°C

Figure 4.10: Isolated cylinder bar exchanging heat with the environment in only one of the
edges

Solution:

Since the lateral of the control volume keeps isolated, the conservation law applied to this
control volume will generate the same energy balance represented by the Equation (4.15),
rewritten below.

∂T k ∂ 2T
= (4.15)
dt ρCp ∂x 2

The initial conditions at t = 0 and x = 0 also keeps the same, but the boundary condition at x =
1m is different this time.At this edge, the heat flowing by conduction from inside the bar to the
edge (at x = 1) is equal to the heat that leaves the bar (at x = 1) by convection (exchanging heat
with the environment), or:

At x = 1m: − kA dT = hA´(T − Tenv ) (4.17)

dx
The Figure 4.11 shows a scheme of the flow of energy at x = 1m.

-kAdT/dx hA´(T-Tenv)

x=

Figure 4.11: Scheme of flow of energy at the edge (at x = 1m)

In which (considering international units):

k = thermal conductivity [J/s m 0C]

A = cross section area from where the energy flows by conduction = πR2 [m2]

dT/dx = temperature gradient [0C/m]

17
 Profa. Liliane Lona – Em revisão

h = global coefficient of heat transfer by convection [J/s m20C]

A´ = area of heat exchange by convection = πR2 [m2]

Tenv = environment temperature [0C]

T = temperature of the bar (in this case, at x = 1m) [0C]

Observe that the cross-section area from where the energy flows by conduction at x = 1 m (A)
is equal to the area from where the bar exchange heat with the environment by convection
(A´), so both terms can be simplified.

It is important to note that the heat represented by both terms of Equation (4.17) has the
same magnitude and flows to the same sense, so both must have the same sign (positive or
negative). The conduction term is positive, because dT is negative, dx is positive and the
Fourier expression brings a minus signal. In this way, the convection term must also be
positive, therefore we used (T-Tenv) to represent the driving force (T >Tenv).

So, the three conditions needed to solve Equation (4.15) are:

At t = 0h, T = 500C, for 0 ≤ L≤ 1m

At x = 0m, T = 700C, for t> 0h

At x = 1m, dT = − h (T − Tenv ) , for t> 0h

dx k

Solving Equation (4.15) with thisnew set of initial and boundary conditionsand assuming h =
300 J/min m20C (details in Chapter 7), the axial profiles of temperature over time shown in
Figure 4.12 can be obtained.

18
 Profa. Liliane Lona – Em revisão

Figure 4.12: Axial profiles of temperature over time when one edge of the bar changes heat
with the environment

One can observe that Figure 4.12 shows temperature profiles very different compared to the
ones presented in Figure 4.6b, because one of the edges exchange heat with the environment.
For the same reason, at steady state a straight line will not be observed anymore to represent
the variation of temperature along the length.

Example 4.6:Now, imagine that the same bar presented in Example 4.5 is not isolate. How
would change the axial profiles of temperature inside the bar until the steady state is reached?

Solution:

For a non-isolated bar, the conservation law applied to the control volume must consider the
heat transfer by convection between the bar and the environment:

Convective term:ℎ × 2𝜋𝜋𝜋𝜋∆𝑥𝑥 × (𝑇𝑇𝑒𝑒𝑒𝑒𝑒𝑒 − 𝑇𝑇) (4.18)

In which (considering international units):

h = the global coefficient of heat transfer by convection [J/s m20C]

2𝜋𝜋𝜋𝜋∆𝑥𝑥 = area from where there is heat exchange by convection between the control volume
and environment [m2]

Tenv = environment temperature [0C]

T = temperature at the control volume [0C]

Adding the expression (4.18) in the energy balance (4.14) used in examples 4.4 and 4.5 (for
isolated bar) we obtain the Equation (4.19). Observe that the energy balance represented
byEquation (4.14) consider the amount of heat that enter, leaves and accumulates in a period
Δt, therefore the expression (4.18) was also multiplied by Δt:

dT dT d dT dT
− kπR 2 ∆t − (−kπR 2 + (−kπR 2 )∆x)∆t + 2hπR∆x(Tenv − T )∆t = ρπR 2 ∆xCp ∆t
dx dx dx dx dt (4.19)

After simplifying terms, Equation (4.19) becomes:

∂T ∂ 2T
ρCp =k + 2 × h (Tenv − T ) (4.20)
dt ∂x 2
R

If Equation (4.20) is written in terms of the diameter of the metal bar (D), and not in terms of
the ration (R), we can obtain after rearranging terms:

𝜕𝜕𝜕𝜕 𝑘𝑘 𝜕𝜕 2 𝑇𝑇 4ℎ
=� � + (𝑇𝑇𝑒𝑒𝑒𝑒𝑒𝑒 − 𝑇𝑇) (4.21)
𝜕𝜕𝜕𝜕 𝜌𝜌𝜌𝜌𝜌𝜌 𝜕𝜕𝑥𝑥 2 𝐷𝐷𝐷𝐷𝐷𝐷𝐷𝐷

19
 Profa. Liliane Lona – Em revisão

The initial/boundary conditions needed to solve Equation (4.21) are the same used in Example
4.5:

At t = 0h, T = 500C, for 0 ≤ L≤ 1m

At x = 0m, T = 700C, for t> 0h

At x = 1m, dT = − h (T − Tenv ) , for t> 0h

dx k

After solving Equation (4.21) (details in Chapter 7), the axial profiles of temperature in the bar
over time are obtained, and shown in Figure 4.13.

Figure 4.13:Axial profiles of temperature over time when the bar changes heat with the
environment.

Compared to the Figure 4.12, we can observe that the temperature of the bar is lower, and
their decay is faster, as expected. The velocity of the decay and the temperaturewill depend on
the numerical values of parameters of the model, such as h, Cp, ρ.

Example 4.7: Imagine a beaker of 1cm of radius (R) and 5cm high (L) open at the top, with only
air inside it.This beaker is placed in an atmosphere containing a certain gas A in a
concentration (CA) equal to 1 mol/m3 (see Figure 4.14). As the concentration of A outside the
beaker is higher (in the beginning, CA inside the beaker is zero, because it contains just air), the
element A will flow from the region with higher (top of the beaker) to the region with lower
(inside the beaker) concentration of A. Let us consider that, for some mechanism chemically
possible, the concentration of A at the bottom of the beaker is always zero. How would be the
axial profile of concentration of A inside the beaker over the time?

20
 Profa. Liliane Lona – Em revisão

Figure 4.14: Beaker initially with air. Diffusion of A from the top to the bottom.

Solution:

The first thing to do to solve this problem is to define the control volume, which, in this case, is
a small cylinder with radius R and length ∆x, as depicted in Figure 4.14. Since we want to know
the axial profile of concentration of A inside the beaker over time, the accumulation term must
be considered. There is no generation and/or consumption of A along the beaker, so the
conservation law applied to the control volume yields: E – S = A.

The flow of A from x to x+∆x occurs by diffusion, due to the difference of concentration of A.
The mass flow by diffusion, according to the Fick´s law is:

dC A
( J A ) x = − DA( )
dx

The amount of A that enter (at x) and leaves (at x+∆x) the control volume is shown as follow:

x (enters) (mol/s) x + ∆ x (leaves) (mol/s)

Massflowbydiffusion dC A dC A d dC A
− DπR 2 − DπR 2 + ( − DπR 2 ) ∆x
dx dx dx dx

In which (in international units)

D = diffusivity of component A in the air (m2/s)

πR2 = area from where the diffusive transport of A takes place (m2)

dCA/dx = concentration gradient for the component A (mol/m3m)

The accumulation of A (in mol) inside the control volume (V) in a period∆t (h) can be obtained
by the difference of A in a time t+∆t and t.

t t + ∆t
Amountof A (mol) V × CA d (V × C A )
V × CA + ∆t
dt

In which

21
 Profa. Liliane Lona – Em revisão

V = control volume (m3) =πR2∆x

The accumulation of A in a time ∆t can be written as:

Accumulation of A in ∆t (in mol) = V dC A ∆t = πR 2 ∆x dC A ∆t

dt dt

The flow of A is given in mol/s and the accumulation term in mol, so the flow of A is multiplied
by ∆t to yield the mass balance shown in Equation (4.22):

dC A 2 dC A d dC A dC A
− DπR 2 ∆t − ( − DπR + ( − DπR 2 ) ∆x) ∆t = πR 2 ∆x ∆t
dx dx dx dx dt (4.22)

Simplifying terms and assuming the diffusion coefficient steady, one can obtain:

∂C A ∂ 2C A
=D
dt ∂x 2 (4.23)

The initial or boundary conditions needed to solve this PDE are:

At t = 0 h, CA = 0 (mol/m3), for 0 ≤ x ≤ 0.05 m

At x = 0 m, CA = 0 (mol/m3), for t> 0 (h)

At x = 0.05m, CA = 1 (mol/m3), for t> 0 (h)

Equation 4.23 can be solved to generate the axial profiles of concentration of A over time, as
shown in Figure 4.7b. The higher the diffusion coefficient the faster the steady state is
reached.

If the axial profile of concentration of A inside the beaker is needed only in the steady state,
the mass balance would yield:

d 2C A
=0
dx 2 (4.24)

With the conditions:

At x = 0 m, CA = 0 mol/m3

At x = 0.05 m, CA = 1 mol/m3

Equation (4.24) can be integrated twice to generate a straight line, as foreseen by Figure 4.7b
to predict the steady state behavior.

Until now, all examples presented in this Chapter 4 consider variation in axial direction. The
example 4.8 will assume changes of temperature in the radial direction. The same procedure
presented so far for developing models can be used. However, it is interesting to pay attention
in the control volume and in the area considered for the flow of energy by diffusion.

22
 Profa. Liliane Lona – Em revisão

Example 4.8: Consider a solid cylinder of copper of length L and radius R1. Assume that the
temperature of the cylinder keeps constant and equal to T0. The copper cylinder is coated with
an annulus made of aluminum, initially at temperature T1. The total radius of the concentric
cylinders (copper plus aluminum) is R2, as depicted in Figure 4.15. The environment
temperature is constant and equal to Tenv. Assume T0 > T1 >Tenv. Although the aluminum ring
exchanges heat with the environment, the two edges of the two concentric cylinders are
isolated. Develop a mathematical model able to predict the radial profiles of temperature in
the aluminumannulus of over time from the beginning (when T = T1) until the steady state is
reached.

Solution:

Initially the annulus is at temperature T1, but its temperature starts changing radially, being
higher near the copper cylinder and lower close to the environment. There is no reason for
axial changes in temperature, because there is no driving force in this direction (edges
isolated).

The first thing to do in order to model this system is define a control volume. In our case, we
can consider a control volume with a small thickness ∆r and length L, as depicted in Figure
4.15.

Figure 4.15: Two concentric cylinders with radial heat conduction along aluminum annulus

The conservation lawapplied to this control volume yields: E – S = A.

As the temperature of the cooper cylinder is higher, heat in the aluminum annulus will flow
from inside to outside. The energy that enters (at r) and leaves (at r + ∆r) the control volume is
due to the conduction of heat (molecule to molecule). These two terms plus the accumulation
of energy in the control volume can be written as follow.

r (enters) r + ∆ r (leaves)

Heat flow rate by dT dT d dT

− kA − kA + ( − kA ) ∆r
conduction (J/h) dr dr dr dr

23
 Profa. Liliane Lona – Em revisão

t t+ ∆t

Amount of energy in ρ × V × Cp × T ρ × V × Cp × T + d ( ρ × V × Cp × T ) ∆t
dt
the control volume (J)

In which (considering the international units):

k = thermal conductivity of the aluminum (J/s m 0C)

ρ= density of the aluminum (kg/m3)

Cp = specific heat of the aluminum (J/kg0C)

dT/dr = temperature gradient in radial direction(0C/m)

T = temperature (0C)

A = cross section area from where the heat flows by diffusion (m2)

V = control volume (m3)

At this point it is very important to understand how the cross-section area (A) and the control
volume (V) are calculated.

The area the energy “sees” when flowing by conduction is 2πrL, which is the superficial area of
the aluminum annulus: the perimeter of the annulus (2πr) times its length (L).

The volume of the control volume is 2πr∆rL, and is calculated considering the area of the base
(2πr∆r) times its length (L). You can be thinking why the area of the base would not be
π(r+∆r)2-πr2. In fact, if you consider that, the area of the base would be 2πr∆rL + π(∆r)2. Since
∆r is very small, the square of the last term is much lower than the first one, and can be
neglected, simplifying the expression for the control volume.

Now we can write the energy balance, multiplying the heat flow by ∆t, tomake the units
compatible:

d ( ρ × 2πr∆rL × Cp × T )
)∆r∆t ) = (4.25)
dT dT d dT
− k 2πrL ∆t − ( − k 2πrL + (−k 2πrL ∆t
dr dr dr dr dt

ENTER – LEAVE = ACUMULATION

Simplifying terms, the energy balance can be rewritten as:

∂ ∂T ∂T (4.26)
k (r ) = rρCp
∂r ∂r ∂t

or

24
 Profa. Liliane Lona – Em revisão

∂ 2 T 1 ∂T ρCp ∂T
+ =
∂r 2 r ∂r k ∂t (4.27)

Observe that the terms 2, π, L, k, Cp and ρ are taken outside the derivative, because they are
constant and do not depend on neither the time nor the radius (we assume that k, Cp and ρ do
not vary with temperature). However, the expression for the cross-section area (2πrL) presents
the independent variable “r” (radius) that must be held inside the partial derivative with
respect to r.

The equation (4.27) needs three initial or boundary conditions: two related to radius and one
related to time. The conditions at t = 0 and at r = R1 are easily obtained, as shown as follow:

t = 0, T = T1 for R1 ≤ r ≤ R2,

t > 0, r = R1 , T = T0 (at this point the annulus of aluminum is in contact with the cylinder of
cooper, which is at T0)

The heat flows by conduction from inside to outside the aluminum annulus until reach the
edgeat r = R2. At thispoint (r = R2), the aluminum annulus exchange heat with the environment,
which is at Tenv. In this way, we can say that at r = R2, the heat that gets to R2 by conduction is
equal to the heat that leaves the annulus (at r = R2) by convection (exchanging heat with the
environment), or:

dT
− kA = hA(T − Tenv )
dr (4.28)

The area A [m2] is the same for both terms of Equation (4.28) (2πR2L) and can be simplified. So,
the three conditions needed to solve Equation (4.27) are:

t = 0, T = T1, for R1 ≤ r ≤ R2,

t > 0, r = R1 , T = T0

t > 0, r = R2, dT = − h (T − Tenv )

dr k

Observe that the convection term does not appear in the energy balance (Equation 4.26 and
4.27), only in the boundary conditions. This is because the control volume considered (see
Figure 4.15) does not interface the environment. This only occurs at r = R2. This is way the
convection term appears only in the boundary condition. A different situation occurred in the
examples 4.1, 4.2, 4.3and 4.6, in which the control volume interfaced the environment, so the
convection term appeared in the energy balanceequation.

The next example will present a more comprehensive situation, in which chemical reactions
also occur.

25
 Profa. Liliane Lona – Em revisão

Example 4.9: Consider a tubular reactor shown in Figure 4.16 in which an irreversible and
exothermic reaction A → B takes place. This reactor, with radius R and length L, has a jacket at
a constant temperature Tj0C, needed to control the reactor temperature. The reactant A is fed
in the reactor in a flow rate Q m3/h, concentration CAin and at temperature Tin0C. The
concentration of A and B and the temperature of the fluid that leaves the reactor are CAf,
CBfand Tf. A chemical engineer needs to build another tubular reactor to produce B in a
different plant. To save money, the engineer wants to know if a shorter reactor could be used
to produce the same amount of B. There is a suspicion that most of A is consumed well before
the end of the reactor. Develop a mathematical model to simulate this reactor in order to
check the viability of building a shorter tube. Assume the system operates in steady state and
that there are no radial or angular profiles of concentration and temperature inside the
reactor.

Solution: The first thing to do in order to model this system is to define a control volume. In
this problem, it is assumed that there are no radial and angular profiles of concentration and
temperature, so all variations occur only along the length. In this way, the control volume will
be a small slice of the reactor, with thickness equal to ∆x, as represented by the Figure 4.16.

Figure 4.16: Tubular reactor to produce the component B.

If there were variationsin concentration and temperature also along the radius, the control
volume would be different. We will revisit this problem in example 4.11, when we will consider
variations in more than one dimension.

The next thing to do to model this reactor is to apply the conservation law to the control
volume, to obtain the mass and balances for the reactor. The jacket temperature keeps
invariable, so there is no need to develop an energy balance to the jacket. The conservation
law applied to the control volume, keeping in mind the reactor operates in steady state, yields:

Mass Balance for the reactant A ⇒ E – L – C = 0

Mass Balance for the product B ⇒ E – L + G = 0

Energy Balance for the reactor ⇒ E – L + G/C = 0

26
 Profa. Liliane Lona – Em revisão

The amount of A that enters the control volume (at x) depends on the flow rate Q (m3/h) and
on the concentration CA (mol/m3), so that the amount of A that goes into the control volume
due to the fluid movement is QCA (mol/h). This is the most important contribution for A
entering the control volume. However, for systems in which the flow rate is very low and/or
for viscousfluids, the diffusion contribution may be important, so in this problem, we will also
consider the axial mass and heat diffusion. Table 4.2 shows the E (entering) and L (leaving)
terms for the mass and energy balance.

Table 4.2: Amounts of A and B and energy that enter and leave the control volume

x (enters) (mol/h) x + ∆ x (leaves) (mol/h)

Flow of A due to the fluid Q CA d (QC A )
QC A + ∆x
movement (mol/h) dx
Flow of A due to the dC A dC A d dC A
− DπR 2 − DπR 2 + ( − DπR 2 ) ∆x
diffusion (mol/h) dx dx dx dx
Flow of B due to the fluid Q CB d (QCB )
QCB + ∆x
movement (mol/h) dx
Flow of B due to the dC B dCB d dCB
− DπR 2 − DπR 2 + ( − DπR 2 ) ∆x
diffusion (mol/h) dx dx dx dx
Flow of Energy due to the 𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 𝑑𝑑(𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇)
𝑄𝑄 𝜌𝜌 𝐶𝐶𝐶𝐶 𝑇𝑇 + ∆𝑥𝑥
fluid movement (J/h) 𝑑𝑑𝑑𝑑
Flow of Energy due to the dT dT d dT
− kπR 2 − kπR 2 + ( − kπR 2 ) ∆x
conduction (J/h) dx dx dx dx

In which:

k = thermal conductivity of the fluid inside the reactor (kJ/h m 0C)

ρ= density of the fluid inside the reactor (kg/m3)

Cp = specific heat of the fluid inside the reactor (kJ/kg0C)

D = diffusivity of reactant A or product B in the fluid inside the reactor (m2/h)

πR2 = area from where the diffusive transport of mass and energy takes place (m2)

Q = volumetric flow rate inside reactor (m3/h)

We will assume that the diffusion coefficient D is the same for reactant A and product B, and
that its value keeps invariable. We will also consider that the fluid that travels the reactor
keeps invariable values for density (𝜌𝜌), specific heat (Cp) and thermal conductivity (k).

Besides the terms shown in Table 4.2, the heat exchanged by convection between reaction
mixture and the jacket must be added in the energy balance.

Heat transfer by convection: U (2πR∆x) (Tj– T) (J/h)

27
 Profa. Liliane Lona – Em revisão

Observe that the exchange area considered in the transfer by convection (2πR∆x) is different
from the cross-section area (πR2) used in the Table 4.2. Do not confuse the areas. Observe also
that there is no mass transfer through the jacked, so no term needs to be added in the mass
balances related to that.

The last terms missing in the mass and energy balances are the consumption and generation
terms, written as follow:

Rate of reaction [kmol/h] ⇒ k´ CA V

Heat of reaction [kJ/h] ⇒ k´ CA V (-∆HR)

In which:

k´ =constant of reaction (1/h)

V = πR2∆x = control volume (m3)

∆HR = heat of reaction (kJ/kmol)

Considering all terms of Table 4.2, plus the terms related to heat convection and reaction, and
afterwards simplifying terms, the mass and energy balances can be written as follow:

dC A d 2C A (4.29)
−Q + DπR 2 − πR 2 k´C A = 0
dx dx 2

dCB d 2C B (4.30)
−Q + DπR 2 + πR 2 k´C A = 0
dx dx 2

dT d 2T (4.31)
− QρCp + kπR 2 + U 2πR (T j − T ) + k´πR 2C A ( − ∆H R ) = 0
dx dx 2

Observe that the Arrhenius constant k´ depends on T, and that the energy balance depends on
CA, so the mass and energy equation must be solved simultaneously.

To solve this system with 3 Ordinary Differential Equations (ODEs), two conditions (initial or
boundary) are needed. In our problem, we know the concentrations and temperature at x = 0
(initial condition):

At x = 0: CA = CAin; CB= CBin = 0; T = Tin

We need another condition to solve the equation system. We can use one of the following
conditions below:

At x = L: CA = CAf; CB = CBf ; T = Tf or

At x →∞, CA = 0; CB = CAin ; T = Tj and dC A = 0 , dC B = 0 , dT = 0

dx dx dx

Observe that if the reactor is very long, all reactant will be consumed and the reaction mixture
will exchange heat with the jacket until its temperature reaches Tj. Observe also the gradient
of concentration reduces along the length as the reactant is consumed, and reaches zero when

28
 Profa. Liliane Lona – Em revisão

there is no reactant available. The same behavior is observed for the gradient of temperature
when the reactant finishes and reactor temperature reaches Tj.

If the diffusion terms were not considered in the mass and energy balance, the model
equations for this reactor would become:

dC A (4.32)
−Q − πR 2 k´C A = 0
dx

dCB (4.33)
−Q + πR 2 k´C A = 0
dx

dT (4.34)
− QρCp + U 2πR (T j − T ) + k´πR 2C A ( − ∆H R ) = 0
dx

That is solved using the initial conditions:

At x = 0: CA = CAin; CB= CBin = 0; T = Tin

Concentration and temperature profiles obtained from solution of Equations 4.29-4.31

(considering diffusional effect) or Equations 4.32-4.34 (not considering the diffusional effects)
can be very similar. This occur because the contributions due to the flow rate (QCA and QρCpT)
are much more important than the diffusion contributions ( − DπR 2 dC A and − kπR 2 dT
dx dx
).The diffusion effects should be considered when a precise model is need to simulate systems
where the flow rate is low and/or the fluid viscosity is high.

The section 4.3 will consider problems where variations occur in more than one spatial
dimension. In this next section, it is important to pay attentionin how the control volume is
defined.

4.3 – Examples considering variation in more than one dimension

In this section, more comprehensive examples will be presented, with variations in more than
one dimension. Let us start with an example considering the axial and radial heat transfer by
conduction in a cylindrical bar. This example is a combination of the concepts presented
previously in Examples 4.6 and 4.8.

Example 4.10: Imagine a cylindrical bar made of a compound with low thermal conductivity.
This bar, with length (L) of 1m and radius (R) equal to 0.3m, is initially at 500C (Figure 4.17). At
some point, one of the two edges of the bar are fixed in a wall that is at 700C (assume the
temperature of the wall does not change with time). Assume the bar exchanges heat with the
environment, which is at 250C. One wants to know the radial and axial profile of temperature
after the steady state is reached. Assume that there is no angular profile of temperature
inside the bar.

29
 Profa. Liliane Lona – Em revisão

Figure 4.17: Cylindrical bar with heat transfer in axial and radial directions

Solution:

One can imagine that the cylinder is at 500C initially, but its temperature close to the wall
starts increasing when it is fixed to the wall, generating an axial profile of temperature. At the
same time, the cylinder starts exchanging heat with the environment, generating also a radial
temperature profile (with higher temperature in the center). Since the radius is big and the
thermal conductivity of the material is low, the radial profile of temperature can be significant,
and must be considered.

To model this system, first we need to define a control volume, small enough to guarantee the
same temperature inside it. Since the temperature changes along x and r, we need to
considered a small ring with length∆x and thickness ∆r as control volume, as shown in Figure
4.17. The control volume for this problem can be visualized as an intersection of the control
volumes drawn in Examples 4.6 (Figure 4.10) and 4.8 (Figure 4.15).

Since one wants to know the temperature profiles after the steady state is reached and since
there is no chemical reaction, the conservation law applied to the control volume shown in
Figure 4.17 yields:

E – L = 0.

The heat enters the control volume at x and r, and leaves it at x + ∆x and r + ∆r. The heat flow
inside the control volume is only by conduction. The convection occurs only at r = R = 0.3m and
x = L = 1m that are regions out of the control volume.

Now using the concept of the infinitesimal variation of the dependent variable with an
independent variable, Table 4.3 and 4.4 can be created.

Table 4.3: Energy that enters and leaves the control volume in the axial direction

x (enters) (J/h) x + ∆ x (leaves) (J/h)

Axial Flow of Energy dT dT d dT
− k 2πr∆r − k 2πr∆r + ( − k 2πr∆r ) ∆x
due to the conduction dx dx dx dx

Table 4.4: Energy that enters and leaves the control volume in the radial direction

30
 Profa. Liliane Lona – Em revisão

r (enters) (J/h) r + ∆ r (leaves) (J/h)

Radial Flow of Energy dT dT d dT
− k 2πr∆x − k 2πr∆x + (− k 2πr∆x )∆r
due to the conduction dr dr dr dr

Observe that the cross-section area from where the energy goes through radial direction is
2πr∆x , that is similar the area found in the Example 4.8, except for the length, that was L and
now is ∆ x. The cross-sectional area from where the energy passes axially is 2πr∆r , that is the
area of the base of the control volume, also calculated in Example 4.8.

Considering the thermal conductivity constant along the cylindrical bar, we can use the terms
of Table 4.3 and 4.4 to obtain the energy balance. After simplifying terms, the energy balance
becomes:

∂ 2T ∂ ∂T
r + (r )=0
∂x 2
∂r ∂r (4.35)

This PDE can be solved using 2 initial conditions (at x = 0 and r = 0) and 2 boundary conditions
(at x = L and at r = R). It is known that the temperature at x = 0 is equal to the temperature of
the wall (700C). At x = L, the heat that comes from diffusion goes out by convection. The same
thing occurs at r = R. One can imagine that there is a symmetrical radial profile of temperature
inside the cylinder, with a maximum temperature in the center. This symmetry conditions
gives us the fourth condition needed to solve the PDE: at r = 0, dT/dr = 0. So, we can write the
following initial/boundary conditions:

At x = 0: T = 700C (4.36)

At x = L: - k 2πr ∆r dT/dx = h 2πr ∆r (T - Tenv) or dT/dx = - h/k (T - Tenv) (4.37)

At r = 0: dT/dr = 0 (4.38)

At r = R: - k 2πr∆x dT/dr = h 2πr ∆x (T - Tenv) or dT/dr = - h/k (T - Tenv) (4.39)

Observe that despite the exchange of heat by convection between the cylinder and the
environment, this convection term does not appear in the energy balance, only in the
boundary conditions. This is because the control volume does not interface the environment.

If one wants to know the axial and radial profiles of temperature at the transient regime, the
accumulation term should be added in the conservation law, as shown in Table 4.5

Table 4.5: Calculus of the accumulation term for the cylindrical bar
t t + ∆t
Amount of energy in V × ρ × Cp × T d (V × ρ × Cp × T )
V × ρ × Cp × T + ∆t
the control volume (J) dt

Keeping in mind the control volume is 2πrΔrΔx, the PDE that represents this system in
transient regime will be:

31
 Profa. Liliane Lona – Em revisão

ρCp ∂T ∂ 2T 1 ∂ ∂T
= + (r )
k ∂t ∂x 2
r ∂r ∂r (4.40)

Besides conditions (4.36) to (3.39), the initial condition related to time needed to solve
Equation 4.40 is:

At t = 0, T = 500C for the entire bar.

Example 4.11:This next example will revisit the Example 4.9, but now we will assume that the
reactor exhibits radial profiles of concentration and temperature. The tubular reactor, where
the reaction A → B takes place, has a jacket that is at constant temperature Tj0C. All operating
conditions and properties are the same of the Example 4.9. One wants to know the axial and
radial profiles of temperature and concentration inside this reactor.

Solution: The main difference between the solution presented in Example 4.9 and the one
presented herein is the definition of the control volume. Since we are considering that both
axial and radial variations are important, we need to consider a small ring with length∆x and
thickness ∆r to represent the control volume, as shown in Figure 4.18.

Figure 4.18: Tubular reactor with changes in axial and radial directions

The conservation law applied to this example generates the same expressions presented
previously (rewritten below). Remember that we are considering steady state (accumulation
term is equal to zero).

Mass Balance for the reactant A ⇒ E – L – C = 0

Mass Balance for the product B ⇒ E – L + G = 0

Energy Balance for the reactor ⇒ E – L + G/C = 0

The amounts of material and energy that enter and leave the control volume can be obtained
using the concept of infinitesimal variation of the dependent variable with the independent
variable.

As done in Example 4.9, we will also consider the axial diffusion of mass and energy. The
amounts of reactant A, product B and energy that enter and leaves the control volume in the

32
 Profa. Liliane Lona – Em revisão

direction x can be seen in Table 4.6. Observe this table is very similar to the Table 4.2,
however the terms are adapted in order to consider the new control volume.

Table 4.6: Amounts of A and B and energy entering and leaving the control volume in the axial
direction

x (enters) (mol/h) x + ∆ x (leaves) (mol/h)

Flow of A due to the fluid Q (2πr∆r/πR2) CA d [Q ( 2πr∆r / πR 2 )C A ]
Q ( 2πr∆r / πR )C A +
2
∆x
movement (mol/h) dx
Flow of A due to the dC A dC A d dC A
− D 2πr∆r − D 2πr∆r + ( − D 2πr∆r ) ∆x
diffusion (mol/h) dx dx dx dx
Flow of B due to the fluid Q (2πr∆r/πR2) CB d [Q ( 2πr∆r / πR 2 )C B ]
Q ( 2πr∆r / πR 2 )C B + ∆x
movement (mol/h) dx
Flow of B due to the dC B dCB d dCB
− D 2πr∆r − D 2πr∆r + ( − D 2πr∆r ) ∆x
diffusion (mol/h) dx dx dx dx
Flow of Energy due to Q ( 2πr∆r / πR 2 ) ρCpT d [Q ( 2πr∆r / πR 2 ) ρCpT ]
Q( 2πr∆r / πR 2 ) ρCpT + ∆x
the fluid movement (J/h) dx

Flow of Energy due to dT dT d dT

− k 2πr∆r − k 2πr∆r + ( − k 2πr∆r ) ∆x
the conduction (J/h) dx dx dx dx

The cross-section area for the diffusion terms in x-direction change from the πR2 to 2πr∆r. The
flow rate Q is related to the total cross-section area πR2. However, the amount of fluid crossing
in the axial direction the cross-section area of the control volume (2πr∆r) is much lower and is
proportional to this new area. In this way, the flow rate crossing axially the new control
volume is Q x [2πr∆r/πR2].

The amounts of reactant A, product B and energy that enter and leaves the control volume in
the radial direction can be seen in Table 4.7. The fluid is fed only in the axial direction, so the
contribution due to the fluid movement in the radial direction is zero. The difference on
concentration and temperature along the radius is due to diffusion. Observe that in the Table
4.7, the cross-section area for the radial diffusion is 2πr∆x.

Table 4.7: Amounts of A and B and energy entering and leaving the control volume in the radial
direction

r (enters) (mol/h) r + ∆ r (leaves) (mol/h)

Flow of A due to the dC A dC A d dC A
− D 2πr∆x − D 2πr∆x + ( − D 2πr∆x ) ∆r
diffusion (mol/h) dr dr dr dr
Flow of B due to the dC B dCB d dC B
− D 2πr∆x − D 2πr∆x + ( − D 2πr∆x ) ∆r
diffusion (mol/h) dr dr dr dr
Flow of Energy due to the dT dT d dT
− k 2πr∆x − k 2πr∆x + ( − k 2πr∆x ) ∆r
conduction (J/h) dr dr dr dr

Observe also that the heat transfer by convection is not considered in the conservation
law,and will appear only in the boundary conditions, because the control volume does not
interface the jacket.

33
 Profa. Liliane Lona – Em revisão

The generation and consumption terms are represented in the same way done in Example 4.9,
however the volume is calculated differently.

Rate of reaction [mol/h] ⇒ k´ CA V

Heat of reaction [J/h] ⇒ k´ CA V (-∆HR)

In which V = 2π∆r∆x = control volume (m3)

All terms of Tables 4.6 and 4.7 plus the reaction terms are combined to generate the system of
Equation that represents this reactor, as follow.

Balance for the reactant A:

Qr ∂C A ∂ 2C A ∂ ∂C A - k´ C = 0 (4.41)
− + rD + D (r ) A
πR ∂x
2
∂x 2
∂r ∂r

Balance for the product B:

Qr ∂CB ∂ 2C B ∂ ∂C B + k´ C = 0 (4.42)
− + rD + D (r ) A
πR ∂x
2
∂x 2
∂r ∂r

Balance of Energy:

QrρCp ∂T ∂ 2T ∂ ∂T + k´ C (-∆H ) = 0 (4.43)

− + kr +k (r ) A R
πR ∂x
2
∂x 2
∂r ∂r

Observe the terms ∆x and ∆r can be simplified. It is assumed that Q, D, ρ, Cp and k are
constant, and can be removed from the derivative.

To solve this equation system, we need four initial or boundary conditions: two in x and two in
r axes. The two conditions in x (at x = 0 and x = L) are the same presented in Example 4.9.

In the radial direction, we can consider the symmetry condition, as occurred in Example 4.8
and 4.10, what give us an initial condition in r = 0. The other condition is given in r = R. For the
energy balance, we can assume that the heat that reaches r=R by conduction leaves the tube
by convection, exchanging heat with the jacket. This give us a boundary condition at r = R, as
per Examples 4.8 and 4.10. For the mass balances, the boundary condition at r=R is based on
the fact that the tube wall is non-permeable for mass transport, so the radial gradients of
concentration at r = R are equal to zero. So, the set of conditions used to solve the system of
PDEs is:

x = 0, for 0 ≤ r ≤ R: CA = CAin; CB= CBin = 0; T = Tin (4.44)

r = 0: for 0 ≤ x ≤ L: dCA/dr = 0; dCB/dr = 0; dT/dr = 0 (4.45)

x = L: for 0 ≤ r ≤ R: CA = CAf; CB = CBf; T = Tf (4.46)

r = R: for 0 ≤ x ≤ L: dCA/dr = 0; dCB/dr = 0; dT/dr = - h/k (T –Tj) (4.47)

34
 Profa. Liliane Lona – Em revisão

Since there are radial profiles of concentration and temperature along allthe reactorlength,
these radial profiles also exist at the exit of the reactor. The concentrations and temperature
values considerate at the boundary condition x = L must be an average of the values over the
radius.

If a mathematical model to predict this reactor also in transient regime is needed, the
accumulation terms must be considered in the conservation law, as calculated in Table 4.8.

Table 4.8: Calculus of accumulation terms for the tubular reactor:

t t + ∆t
Amountof A (mol) 2πr∆r∆x × C A d (2πr∆r∆x × C A )
2πr∆r∆x × C A + ∆t
dt
Amountof B (mol) 2πr∆r∆x × CB d (2πr∆r∆x × CB )
2πr∆r∆x × CB + ∆t
dt
Amount of energy in 2πr∆r∆x × ρ × Cp × T 2πr∆r∆x × ρ × Cp × T + d (2πr∆r∆x × ρ × Cp × T ) ∆t
the control volume (J) dt

The model to represent the reactor in transient regime can be represented by Equations (4.48)
– (4.50) (compare to model in steady state regime, equations 4.41 – 4.43).

Balance for the reactant A:

∂C A Qr ∂C A ∂ 2C A ∂ ∂C A - k´ C (4.48)
r =− + rD + D (r ) A
∂t πR ∂x
2
∂x 2 ∂r ∂r

Balance for the product B:

∂C B Qr ∂CB ∂ 2C B ∂ ∂C B + k´ C (4.49)
r = − + rD + D (r ) A
∂t πR ∂x
2
∂x 2 ∂r ∂r

Balance of Energy:

∂T QrρCp ∂T ∂ 2T ∂ ∂T + k´ C (-∆H )
rρCp =− + kr +k (r ) A R (4.50)
∂t πR 2
∂x ∂x 2 ∂r ∂r

The initial/boundary conditions in r and x are the same presented previously (Equation 4.44 to
4.47), and the initial condition for the time needed to solve the transient model is:

At t = 0: CA = CAo; CB= CBo; T = To (4.51)

The initial conditions at t = 0 (CAo,CBo, and To) can be a constant or can vary with x and/or r
(function of x and/or r).

Finding boundaries conditions to solve differential equations generated from the modeling can
be very tricky. Besides, depending on the boundaries conditions used, the numerical solution
of the models can become less or more complicated. Some works in literature, like Boyadjiev
(2010), show examples applying boundary conditions in different ways.

35
 Profa. Liliane Lona – Em revisão

Chapters 3 and 4 bring enough information to develop models for piece of equipment usedin
chemical plants in different levels of complexity. Depending on the complexity of the system
being modeled, different kinds of equations are generated. It is possible to obtain algebraic
equations, ODEs (of first and second degree), and PDE (varying with 2, 3 or 4 independent
variables). The second part of this book will focus on the numerical solution for the equations
generated from the modeling. Chapter 5, 6 and 7 will solve systems represented by algebraic
equations, ODE and PDE respectively, and will use Excel as computational tool.

Proposed problems

4.1 )Imagine a very thin disc of thickness δinitially at 500C. This disc has a small orifice in the
middle, in which a pin with temperature 1000C is engaged in. The environment is at 250C.
Assume that both the environment and the pin temperatures do not change over time.
Assume also there is no temperature profile along the thickness δ of the disc.

(a) Make a sketch of how the disc temperature would vary along the radius and over time until
the steady state is reached. Draw both curves in the same graph.

(b) Do the modeling of this system and find the partial differential equation to represent how
the temperature changes along the radius and over time.

(c) Define all initial/boundary conditions necessary to solve this PDE.

4.2 ) Consider the heat conduction in an extended surfaces represented by the fin shown in
the Figure below, with dimension L x L x 10L. The fin, which was initially at 250C, is fixed in a
wall at 1500C. The fin exchange heat with the environment that is at 250C. Assume the wall and
environment temperatures do not change over time.

(a) Find the partial differential equation the represents the fin temperature along x, y and z,
and over time.

(b) Find the initial/boundary necessary to solve the PDE.

4.3 ) Imagine a solid cone as depicted below.The lower and the upper bases of the cone are 2
and 12 cm respectively and its height is 5 cm. This cone was initially at a 350C and the

36
 Profa. Liliane Lona – Em revisão

environment temperature (Tenv) is 250C. The lower base of the cone is submitted to a
temperature Tb equal to 1800C. The lateral of the cone is isolated, and the cone exchange heat
with the environment only through the upper base. Assume there is no radial and angular
profiles of temperature in the cone and that Tb and Tenv do not change with time.

(a) Define the ODE that represents the variation of temperature along the height of the solid
cone at steady state. (Tip: represent the radius of the cone as function of the height to obtain
the ODE).

(b) Define the initial/boundary conditions needed to solve the ODE

4.4) Imagine two tubular reactors with radius R1 (m) and R2 (m), setin a consecutive way, in
which the exothermic reaction A→B takes place. This system exchanges heat with the
environment that is at Tenv (0C). The radius of the first reactor is three times the radius of the
second (R1 = 3R2). The length of the first and the second reactor are L1 and L2 (m) respectively.
Reactant A is fed in the reaction system at a concentration CAin(mol/m3), temperature Tin (0C)
and in a flow rate Q (m3/h),as depicted by the figure below. There is a sample collection
between the two reactors, so it is possible to know the concentrations and temperature of the
reaction mixture at this point.

Assume the reactors operate in steady state. Consider thatall properties of the fluid (density,
specific heat, thermal conductivity, mass diffusivity) are the same for both reactors and do not
vary along the length. The global coefficient of heat transfer by convention for this reaction
system is h [J/h m2 0C].

(a) Draw the control volume and develop the energy and the mass balance for the reactant A
for the first reactor. In this case, consider mass and energy diffusion only in axial direction.
Define all initial/boundary conditions to solve the model equations.
(b) Draw the control volume and develop the energy and the mass balance for the reactant A
for the second reactor. In this case, consider the mass and energy diffusion only in radial
direction. Define all initial/boundary conditions to solve the model equations.

37
 Profa. Liliane Lona – Em revisão

4.5 )Imagine a solid sphere with radius Rs hanged by a very thin string in the center of a closed
cylindrical recipient full of a fluid initially at 500 C, as depicted in the figure below. The air that
surrounds the cylinder is at 250C and its temperature keeps invariable over time. The solid
sphere was at 1000C before being immersed in the cylinder.It can be assumed that all
properties for the fluid and for the sphere do not change over time. Assume the temperature
of the fluid is the same along the entire cylinder.

a) Consider the sphere is made of a very conductive material and that its radius Rs is very
small, in a way there is no temperature profiles inside the sphere. Develop an energy
balance for the sphere and for the fluid and obtain the two differential equations that
represent this system over time. Define all initial/boundary conditions.
b) Now consider the sphere is made of a material with very low thermal conductivity and its
radius Rs is big enough to become significant the radialtemperature profiles insideit.
Develop again the energy balances for this new situation and define all initial/boundary
conditions. The energy balance for the fluid is different in this case?

For cases a and b make assumptions for the models and define parameters if necessary.
Remember the area and volume of a sphere is 4𝜋𝜋𝑅𝑅 2 and(4/3)𝜋𝜋𝑅𝑅 3 respectively.

4.6) Imagine a tubular reactor in which the irreversible reaction A→B takes place. Reactant A is
fed the reactor in a flow rate Q (m3/h) and in a concentration (CAin) 1mol/L. The reactor has a
jacket in which a thermal fluid at temperature Tj is fed in a concurrent way with the reactional
mixture. However, the jacket temperature changes along the reactor length, because the
jacket exchanges heat with the reactional mixture and with the environment, which is at a
constant temperature Tenv = 250C. Assume there is no radial and angular profiles inside the
reactor and that it operates at steady state. Assume also the diffusional effects in axial
direction are not important. Find the differential equations system that represents the
variation of concentration of reactant A, temperature of the reaction mixture andtemperature
of the jacket along the length. Find the initial/boundary conditions needed to solve the
equations system. Define parameters and hypothesis for your model when needed.

4.7 ) Imagine that a solid cylinder with radius and length of 2m and 10m respectively, initially
at 300C, is fixed between two surfaces at constant temperatures (1000C at the right and 200C at
the left).

a) Assuming there is no heat exchange with the environment (isolated cylinder), do a sketch of
how would be the axial profiles of temperature in the cylinder until the steady state is reached.

38
 Profa. Liliane Lona – Em revisão

b) Assuming thermal conductivity of the material equal to 1 J/h m °C, what is the temperature
and the energy flux at the positions x = 2m, 5m and 9m at the steady state?

REFERENCES

Kern, Q.D., Process Heat Transfer, McGraw-Hill Book Company, (1950)

Boyadjiev, C., Theoretical Chemical Engineering Modeling and Simulation, Springer, 2010

Bergman, T. L., Lavine, A. S., Incropera, F. P. and Dewitt, D. P., Fundamentals of heat and mass
transfer, 7th ed., John Wiley & Sons, Inc. (2011)

Bird, R. B., Steward, W. E. and Lightfoot, E. N., Transport Phenomena, 2nd ed., John Wiley &
Sons, Inc. (2007)

Fox, R. W., McDonald, A. T., Pritchard, P. J., Leylegian, J. C., Fluid Mechanics, 8th ed., John Wiley
& Sons, Inc. (2012)

Welty, J. R., Wicks, C. E., Wilson, R. E. and Rorrer, G. L., Fundamentals of Momentum, Heat and
Mass Transfer, 5th ed., John Wiley & Sons, Inc. (2007)

39