Summer 2004

Cross Flow Heat Exchanger Experiment

1. To understand the basic operation of heat exchangers.

2. To demonstrate the basic equations of heat exchanger operation.

BACKGROUND

A heat exchanger is a heat transfer device whose purpose is the transfer of energy from
one moving fluid stream to another moving fluid stream. It is the most common of
heat transfer devices and examples include your car radiator and the condenser units on
air conditioning systems. The overall energy transfer is dictated by thermodynamics
and the First Law. To perform the thermodynamic analysis on a heat exchanger, we
consider the control volume shown below.

Hot Stream

Qint
Cold Stream

c.v.

W = 0, Q = 0

Note that although there is heat transfer from the hot fluid stream to the cold fluid
stream, there is no work or heat transfer from the control volume (c.v.) to the
surrounds. The first law for this control volume is then written as

(1)

Considering that we have two flows into the control volume and two flows out of the
control volume, we may write a more specific form of the first law as

(2)

1
 Summer 2004

or rearranging by grouping the streams

(3)

This, then, is the most general form of the First Law for a heat exchanger. However,
for many heat exchangers there is not a phase change occurring for either fluid stream
and the fluids are either incompressible liquids or ideal gases. Under these two
conditions, we may represent the enthalpies in terms of temperature (a much more
measurable quantity) by using the appropriate equation of state ( ), which will
introduce the specific heat). Then our First Law becomes in final form

(4)

Recall that in this transformation from enthalpies to temperatures, we have assumed

Unfortunately, thermodynamics does not tell the whole story of a heat exchanger's
performance. To achieve the energy transfer predicted by the First Law the principles
of convection and conduction heat transfer must be applied. To apply these principles
we consider a very small length of the heat exchanger, x, as shown below.

1
C o n v e c t io n h H PH x
H ot
F lu id
W a ll C o n d u c t io n R w
1
C o ld
C o n v e c t io n h C PC  x
F lu id

We note that the following heat transfer processes are at work. First, there is
convective heat transfer from the hot fluid to the wall surface, next there is conduction
through the wall, and finally there is convection from the wall surface into the cold
fluid. This series of heat transfer process is ideally modeled by the thermal circuit
model, which is shown in the above figure. The total thermal resistance is then given
as

2
 Summer 2004

(5)

Utilizing this, our heat transfer between the two fluid streams over this small length
segment x is

(6)

Introducing the concept of an overall heat transfer coefficient, U, so that U times the
heat transfer surface area is equal to the thermal conductance (one over the thermal
resistance), we write

(7)

where P is the perimeter such that Pdx is the differential heat transfer surface area
(A). To obtain the total heat transfer between the two fluids inside the heat
exchanger, the above expression is integrated from 0 to L (the length of the heat
exchanger),

(8)

which from our thermodynamics is also equal to

(9)

We now have a relationship between the heat transfer and thermodynamics. The
difficulty with utilizing Eq. (7) lies in evaluating the integral. In order to evaluate the
integral, we must know the functional forms of the temperatures, T H and TC. The only
way to do this is to write the appropriate differential energy equation for both fluid
streams and solve these coupled equations for the temperatures. Unfortunately, the
flow in most heat exchangers is so complicated that a simple solution to the differential
equation is not possible and we are forced to take another approach.

The approach that we will be using is based upon the dynamic scaling and
dimensionless parameter work you saw in your fluid mechanics course. We begin with
some definitions:

3
 Summer 2004

Flow Heat Capacity

Effectiveness (11)

Number of Transfer Units (12)

Our next step would be to employ dynamic similarity to obtain a relationship among
our three dimensionless parameters, CR, , and NTU. We can partially show this by
beginning with Eq. (7). We modify Eq. (7) by noting that by definition

(13)

where Tavg is the average temperature difference between the hot and cold fluids as
they pass through the heat exchanger. Then our heat transfer is given by

(14)

We still do not know Tavg, but let us assume that it depends linearly on the maximum
possible temperature difference, , and that the constant or proportionality is
really a function of UA, Cmin, and Cmax. Then we may write

(15)

We would now like to normalize this heat flow, bound it between zero and one, which
we can do by dividing Eq. (15) by the maximum possible heat transfer (which will give
us the effectiveness) to obtain

4
 Summer 2004

(16)

which after simplification can be rewritten

(17)

We recognize UA/Cmin as the NTU and that the function can be written equivalently in
terms of NTU and CR, rather than the three parameters stated. Then we have

(18)

but since NTU appears in the function, it is redundant to have it out in front, so that we
may finally write

(19)

This is the basis for one of the most powerful tools in heat exchanger analysis, the
effectiveness-NTU approach. In your heat transfer text book you will find these
effectiveness-NTU relationships for a variety of heat exchangers in both equation form
and graphically. A typical graphical relationship is shown below for a crossflow heat
exchanger where both fluids are unmixed.

5
 Summer 2004

Figure 3. Effectiveness - NTU Relationship for Counterflow Heat Exchanger

0.9

0.8

0.7

0.6
Effectiveness

0.5

Cr = 1.0
0.4 Cr = 0.75
Cr = 0.5
Cr = 0.25
0.3
Cr = 0.0

0.2

0.1

0.0
0.0 1.0 2.0 3.0 4.0 5.0
NTU

A crossflow heat exchanger is one in which the two fluid streams are perpendicular to
each other. The unmixed condition relates to the absence of any significant fluid
velocity components in the direction normal to the streamwise direction. For example,
a fluid flowing in a circular pipe is unmixed since there are no velocity components in
the r or  directions, which are the directions normal to the streamwise direction which
is z. The above graph may also be represented by an equation as

(20)

6
 Summer 2004

A final note about this equation and its corresponding graph concerns effectiveness
behavior when the NTU is small. When the NTU is less that 0.5, all of the C R curves
collapse. Since the graph has a C R = 0 curve, one could take the effectiveness values at
CR=0 to be valid for all C R's when NTU is small. However, if we look at Eq. (20)
carefully we see that for CR = 0, we would be dividing by zero. Hence, l'Hopital's rule
would have to be applied which yields a much simpler equation for cases when C R = 0
or NTU < 0.5 of the form

(21)

EXPERIMENTAL BACKGROUND

In this experiment the student will utilize the cross flow heat exchanger apparatus in
the Heat Transfer Lab as the heat exchanger to be studied. As shown below, the hot
fluid flows inside a tube, while the cold fluid is directed to flow perpendicular to the
outside of this tube.

Cold Fluid

Hot Fluid

Using an effectiveness - NTU approach an experimental value of the NTU, NTU exp,
will be determined. Using a thermal circuit, calculation with the appropriate Nusselt
number correlations for the calculation of the convective heat transfer coefficients, a
theoretical NTU, NTUth, will be determined. The basic goal of this experiment is to
compare NTUexp to NTUth. By taking temperature and flow measurements an
experimental value of the effectiveness may be calculated from,

7
 Summer 2004

(22)

For this experiment, these values will be quite small which indicate that the NTU exp
will be small and that Eq. (21) can be used to calculate NTU exp. Using a thermal circuit
model an expression for the UA product can be obtained as

(23)

The convective heat transfer coefficients are calculated from the appropriate Nusselt
number correlations. The hot fluid heat transfer coefficient is calculated for turbulent
flow inside a circular channel, while the cold fluid heat transfer coefficient is
calculated for cross flow over a cylinder. Using the definition of NTU, Eq. (12), NTU th
can be determined.

During the experiment the cold water flow rate is varied by the student from trial to
trial. This will affect the operating temperatures and, consequently, the effectiveness
and NTUexp. The varying of the cold water flow rate will also influence the cold water
convective heat transfer coefficient and, through propagation in the calculations, the
NTUth. We see this influence of cold water flow rate in the sketch below.
 c  T c,out , T H,out    NTU exp
m
m


 c  v C  ReC  Nu C  hC  UA  NTUth
m

Hence, NTUth can be identified with a specific value of the effectiveness. So the final
result, as shown below, will be a graph of effectiveness versus NTU with curves of
both NTUexp and NTUth.

8
 Summer 2004

NTUth
NTUexp

Number of Transfer Units

Effectiveness

The key uncertainties to be considered are those associated with the effectiveness and
NTUexp. If we assume that C H = Cmin, then the effectiveness is calculated from the
temperature measurements by

(24)

Thus we have

(25)

so that the uncertainty in  is given by

(26)

The partial derivatives can be easily obtained from Eq. (24). For NTU exp, we have

(27)

and since NTUexp is only a function of 

9
 Summer 2004

(28)

where d is found from above.

PROCEDURE

1. Make the appropriate dimension measurements on the heat exchanger that appear in
the UA product expression.

2. With the help of your lab instructor turn on the system.

3. Turn both the hot water supply valve and the cold water supply valve to full open.
Note the maximum flow rate for the cold water. Since six trials are required, the
student can estimate the appropriate cold water flow rates for the trials.

3. Set the cold water supply valve so as to provide the appropriate cold water flow rate
for the trial. Allow the system to come to steady state and record inlet and outlet
temperatures of the cold and hot water and the hot and cold water flow rates.

4. Repeat the experiment for at least six different flow rates cold water.

DATA ANALYSIS

1. Each trial should be recorded on the data spread sheets available in the lab. A
template of the spreadsheet is attached. The spreadsheet performs the calculations
noted in items 2-4. However, the student is required to submit a sample calculation
of the effectiveness and NTUexp.

2. From the experimental data calculate the effectiveness for each trial. Using the
experimental effectiveness and Eq. (21), calculate NTU exp.

3. Using a standard correlation from your textbook, calculate the hot water and cold
water heat transfer coefficients.

4. Using a thermal circuit model calculate the UA product. Using the definition of
NTU determine NTUth.

5. Calculate the uncertainty in the effectiveness and the experimental NTU.

6. Plot the effectiveness versus the NTU for both NTU exp and NTUth. Show the
experimental uncertainties using error bars.

10
 Summer 2004

SUGGESTIONS FOR DISCUSSION

1. How do NTUexp and NTUth compare?

2. What errors may be present in your experimental analysis?

3. Is the cold water flow rate truly an external flow?

11