Sie sind auf Seite 1von 25

An Empirical Proof for Turbulent Flow in a Heat

Exchanger Using Nusselt Number

Didi He
Betty Huang
Angel Morales
James Wu
Transport Laboratory (06-363)
Experiment 2 Lab Report
Professor Tilton and Matt Cline
May 14, 2001

2
Table of Contents
An Empirical Proof for Turbulent Flow in a Heat Exchanger Using Nusselt Number......1
List of Figures....................................................................................................................ii
List of Equations...............................................................................................................iii
Abstract............................................................................................................................iv
1 INTRODUCTION............................................................................................................1
2 BACKGROUND & THEORY...........................................................................................1
2.1 Heat Exchangers......................................................................................................1
2.2 Thermodynamics Behind Heat Exchangers............................................................1
2.3 Correlations between v, NNu, NRe, and NPr..........................................................2
3 EXPERIMENTAL SETUP...............................................................................................4
3.1 Equipment................................................................................................................4
3.2 Procedure.................................................................................................................4
4 RESULTS.......................................................................................................................4
4.1 Operating Space of Heat Exchanger.......................................................................4
4.2 Correlations between U and Nusselt Number (NNu)...............................................5
4.2.1 Tube Side Correlations......................................................................................5
4.2.2 Shell Side Correlations......................................................................................6
5 DISCUSSION..................................................................................................................7
5.1 Analysis of Results...................................................................................................7
5.2 Sensitivity Analysis for Assumptions Made.............................................................8
5.3 Sources of Error.......................................................................................................8
6 CONCLUSIONS & RECOMMENDATIONS...................................................................9
6.1 Conclusions..............................................................................................................9
6.2 Future Lab Work.......................................................................................................9
7 Nomenclature................................................................................................................10
8 References & Sources Cited........................................................................................11

i
List of Figures

Figure 2.1.1: 2-4 Heat Exchanger......................................................................................1


Figure 4.1.2: Operating Space for Primary Heat Exchanger.............................................5
Figure 4.2.3.1 Double Reciprocal Plot of U vs. NNu (Tube Side).....................................6
Figure 4.2.2.4: Double Reciprocal Plot of Nusselt Number and U ...................................7
Figure A.2.A.5: Data with Non-Steady-State Values.........................................................2
Figure A.2.A.6: Data Used in Data Analysis (only steady state values)............................3
Figure C.7: Wilson Plot: Constant HW Flow (Tube Side)..................................................1
Figure C.8: Wilson Plot: Constant CW Flow (Shell Side, square).....................................2
Figure C.9: Modified Wilson Plot: Constant CW Flow (Shell Side, Triangular).................3

ii
List of Equations

Equation 2.2.1: Overall Heat..............................................................................................1


Equation 2.2.2: Logarithmic Mean Temperature...............................................................2
Equation 2.3.3: Nusselt Number........................................................................................2
Equation 2.3.4: Reynolds Number Equation.....................................................................2
Equation 2.3.5: Nusselt Number (Expanded)....................................................................3
Equation 2.3.6: Nusselt Number for Shell Side Flow........................................................3
Equation 2.3.7: Reynolds Number for Shell Side Flow.....................................................3
Equation 2.3.8: Definition of ν max...................................................................................3

iii
Abstract
In this experiment, we used a 2-4 pass shell-and-tube heat exchanger to determine the
relationship between overall heat transfer coefficient (U) and Nusselt number (NNu).
From the Wilson Method (Bruce, Wilson Method, 2000), we know that a double
reciprocal plot of 1/U versus 1/v0.8, where v is fluid velocity, results in a linear
relationship. Since v and the Reynolds number (NRe) are the only significantly changing
parameter in the NRe and the NNu, respectively, we are able to directly relate overall heat
transfer coefficient to the Nusselt number. Furthermore, plotting 1/U versus 1/NNu also
results in a linear relationship. The SSCF 4-pass shell-and-tube heat exchanger has an
operating space with an inlet hot water temperature range from 60°F to 200°F with the
temperature most often between 100°F and 150°F. The hot and cold water flow rates
ranged from 1 gallon per minute (gpm) to 15 gpm.

iv
1 INTRODUCTION
The Chemical Engineering Department at Carnegie Mellon University commissioned
this experiment under the guidance of Professor Robert Tilton and Mr. Matt Cline. Our
goal was to test the correlation between the overall heat transfer coefficient (U) and the
Nusselt number (NNu) in a counter-current shell-and-tube heat exchanger. Since the
heat exchanger is new, no data is available from previous groups. However, a similar
experiment, (Bruce, Wilson Method, 2000), showed a linear relationship between 1/U
vs. 1/v0.8 where v is fluid velocity. This relationship is an empirically determined
equation for NNu.

2 BACKGROUND & THEORY


2.1 Heat Exchangers
A heat exchanger is an apparatus used for the exchange of heat from one liquid to
another. The heat exchanger does not mix the fluids together in order to achieve heat
transfer; heat transfer occurs through a medium (i.e. the pipe wall of the heat
exchanger). One type of heat exchanger is a shell and tube heat exchanger (Figure
2.1.1). The process we looked at is the cooling of hot water. Hot water flows around
the tubes; cold water flows inside the tubes. In a shell-and-tube heat exchanger, the
flow of the two liquids is counter-current.

Prieve, D. 2001

Figure 2.1.1: 2-4 Heat Exchanger

2.2 Thermodynamics Behind Heat Exchangers


The basic principle a heat exchanger works on is the first law of thermodynamics.
Equation 2.2.1 shows the amount heat transferred in a heat exchanger.

Equation 2.2.1: Overall Heat

Q = UA T ∆T lm

1
AT , U, and ∆Tlm are the total area of the heat exchanger, the overall heat transfer
coefficient, and the logarithmic mean temperature difference of the heat exchanger,
respectively.

We used a logarithmic mean temperature because there are two temperature inlets and
two temperature outlets. Equation 2.2.2 shows the calculation for logarithmic mean
temperature.

Equation 2.2.2: Logarithmic Mean Temperature

' '
( ∆T 2−∆T 1)
∆T lm =
∆T ' 2
ln( '
)
∆T 1

∆ T’1 = T’1 - T1 where T’1 is the inlet temperature of the hot fluid, and T1 is the inlet
temperature of the cold fluid. T’2 = T’2 - T2 where T’2 and T2 are the outlet temperature
differences of the hot fluid and the cold fluid, respectively.

2.3 Correlations between v, NNu, NRe, and NPr


In Equation 2.3.1, we related Nusselt number to Reynolds number, another measure of
fluid behavior.

Equation 2.3.3: Nusselt Number

hD µ 1
N Nu = = .027 N Re .8N Pr 3 ( b ) 0.14
k µw

NNu, h, D, k, NRe, NPr, µ b, and µ w are Nusselt number, convective heat transfer
coefficient, tube inner diameter, conductive heat transfer coefficient, Reynolds number,
Prandtl number, bulk viscosity, and viscosity of water, respectively.

This equation is only valid for turbulent flow which is defined as any fluid flow with a
Reynolds number of greater than 6000. Velocity, used to calculate Reynolds number, is
an important factor when calculating the Nusselt number.

Equation 2.3.4: Reynolds Number Equation

Dv ρ
N Re =
µ

D, v, µ , and ρ are tube inner diameter, velocity, viscosity, and density, respectively.

By substituting Equation 2.3.2 into Equation 2.3.1, we obtain a relationship between


Nusselt number and fluid velocity in the tube. (Equation 2.3.3).

2
Equation 2.3.5: Nusselt Number (Expanded)

 Dv ρ .8 µb 0.14 1
N Nu = .027 
 µ   N Pr ( µ )
3

  w

To calculate NNu on the shell side, we used Equation 2.3.4 (Geankoplis, 252).

Equation 2.3.6: Nusselt Number for Shell Side Flow

1
N Nu ,Shell = C ( N Re, Shell ) m
N Pr 3

NNu,Shell, C, NRe,Shell, and m are Nusselt number (shell side), tube arrangement constant
(Geankoplis, Table 4.6-2), Reynolds number (shell side), and tube arrangement constant
(Geankoplis, Table 4.6-2), respectively.

Equation 2.3.5 defines the Reynolds Number for shell side flow.

Equation 2.3.7: Reynolds Number for Shell Side Flow

D ⋅ν max ⋅ ρ
N Re, Shell =
µ

ν max is the maximum velocity for shell side flow.

Equation 2.3.6 show how vmax for shell side flow is defined.

Equation 2.3.8: Definition of ν max

ν ⋅S
n
ν max
S −D
n
Sn is the pitch distance, the distance from the center of one tube to the center of the next tube.
.

From Table 4.6-2 in Geankoplis, we can determine constants C and m using the pitch
distance between the tubes and the diameter of the tubes. Pitch distance will vary
depending on whether the tube arrangement is in-line (square) or staggered (triangular).

3
3 EXPERIMENTAL SETUP
3.1 Equipment
There were three major components to our experiment: a primary heat exchanger,
secondary heat exchanger, and LabView. The primary heat exchanger is the 2-4-pass
shell and tube liquid-liquid exchange. The secondary heat exchanger is the steam-
liquid exchanger, which is connected to the primary heat exchanger as a source for the
hot water (shell side). The third component is our operating program, LabView.
LabView sets the hot and cold water flow rates in gallons per minute. It also monitors
the cold and hot inlet and outlet temperatures. The complete list of data components
collected by LabView is in Appendix A.1. The water source was the city water main
line.

3.2 Procedure
In this experiment, there were four main steps.

1. Operating Boundaries. We determined the operating boundaries of the heat


exchanger. We ran the heat exchanger at various combinations of cold and hot
water flow rates (1 gpm to 15 gpm). We determined an operating space by
plotting hot water flow rates versus hot water temperature (Figure 4.1.1).

2. Tube Side Analysis. We ran trials keeping the hot water flow rate constant and
varying the cold water flow rates. After each flow rate increase, we waited until
the flow rates reached steady state.

3. Shell Side Analysis. We repeated step 2 except we kept the cold water flow rate
constant and varied the hot water flow rate.

4. Data Duplication. We repeated steps 2 and 3 a few times to ensure our data
was reproducible.

We used LabView to record all the data.

4 RESULTS
4.1 Operating Space of Heat Exchanger
Figure 4.1.1 shows the operating space of the primary heat exchanger; Figure 4.1.1
shows the relationship between hot water flow rate and hot water inlet temperature.
The SSCF 4-pass shell-and-tube heat exchanger has an operating space with an inlet
hot water temperature range from 60°F to 200°F. As shown in Figure 4.1.1, the
temperature was most often between 100°F and 150°F.

4
Figure 4.1.2: Operating Space for Primary Heat Exchanger

4.2 Correlations between U and Nusselt Number (NNu)


The Wilson plot relates U to fluid velocity by a double reciprocal plot. 1/U versus 1/v0.8
displays a linear relationship for fully developed turbulent flows (Bruce, Wilson Method,
2000). Since NNu is primarily dependent on v, we can derive a correlation between U
and NNu.

4.2.1 Tube Side Correlations


For our SSCF 4-pass shell-and-tube heat exchanger, fluid velocity inside the tube
ranged from 289 ft/min to 784 ft/min. Using Equation 2.2.1 to calculate U, we obtained
a linear relationship between 1/U and 1/v0.8 (Appendix C). The Reynolds number
ranged from 7700 to 19500.

Based on the Wilson plot, we plotted a double reciprocal plot of U and N Nu. For the
operating temperature range of 100°F to 200°F, the changes in Prandtl number and
viscosity were negligible. From the Sieder-Tate equation, we determined that the only
significantly changing parameter of NRe is the fluid velocity. Therefore, the relationship
between 1/U and 1/NNu is also linear (Figure 4.2.1.1).

5
Figure 4.2.3.1 Double Reciprocal Plot of U vs. NNu (Tube Side)

4.2.2 Shell Side Correlations


The graph of U versus velocity for the shell side displays a negative slope because the
heat transfer is from the shell side to the tube side. The flow on the shell side is
turbulent; the shell side Reynolds number ranges from 41,000 to 124,000 for staggered
arrangements, and from 16,000 to 48,000 for in-line arrangements.

Figure C.2.1 shows the linear relationship between the overall heat transfer coefficient
and the velocity of hot water on the shell side (square pitch). Since the shell side flow is
very turbulent, the Wilson plot shows a linear relationship between 1/U and 1/v0.8. We
calculated the shell side Nusselt number using Equation 2.3.4. Since the N Nu,Shell uses
NRe,Shell, we modify the Wilson Plot for the shell side to be 1/U vs. 1/v^m where m is
0.632 (square pitch), the tube arrangement constant obtained from Table 4.6-2 in
Geankoplis.

Similar to tube side Nusselt number plot (Fig. 4.2.1.2), we also see a linear relationship
for a double reciprocal plot of U and NNu for the shell side (Figure 4.2.2.1).

6
Figure 4.2.2.4: Double Reciprocal Plot of Nusselt Number and U
(Shell Side: square pitch)

5 DISCUSSION
5.1 Analysis of Results
As we predicted, we found a linear relationship between the Nusselt number and the
overall heat transfer coefficient. We observed a linear relationship between U and v as
predicted by the Wilson plot (Bruce, Wilson Method, 2000) with our experimental data
(Appendix C). By relating velocity to Nusselt number with Reynolds number, we show
that U is related to Nusselt number.

We discarded data points that are not in the turbulent regions. Since we based our
calculations on formulas for turbulent flow, we only analyzed the data points that were
clearly in turbulent flow regions.

The shell side flow is turbulent due to the minimum pitch distance assumption. We
made this assumption because "designers prefer to employ the minimum recommended
tube pitch.” (Mukherjee,1998). Therefore, 5/4 of the tube diameter is used for the
triangular pitch, and an additional quarter inch is added for the square pitch (Mukherjee,
1998). Using a tube diameter of .25 inches, we obtained a triangular pitch distance of
0.313 inches, and a square pitch distance of 0.5 inches. Appendix B.2 contains detailed
calculations.

7
Equation 2.3.4 is only valid for more than heat exchangers with at least 10 rows of
tubes. In our heat exchanger, we calculated 6.7 rows for triangular pitch and 4.2 rows
for square pitch; therefore, we needed to determine correction factors for the equations.
Since the distance between baffles cannot exceed the diameter of the heat exchanger,
the number of rows traversed should be round down. Using the triangular pitch, six
rows of tubes were traversed. Using the square pitch, four rows of tubes were
traversed. Triangular pitch is a staggered formation; square pitch is an in-line formation.
According to Table 4.6-3 in Geankoplis 1993, the correction factors are .94 and .90 for
triangular and square pitch, respectively (Detailed calculations shown in Appendix B.2).

5.2 Sensitivity Analysis for Assumptions Made


We made three assumptions in our calculations; they are:
1. Tube side thickness is 1/32 inches,
2. Shell inner diameter is two inches,
3. Total tube cross-sectional area should be quartered since the heat exchanger is
four-pass, and
4. Tube pitch is the minimum allowed by TEMA standards.

The values we assumed are in Appendix B.2.

In order to determine how greatly these assumptions can affect our calculations, we can
change all four assumptions simultaneously to either their individual upper or lower
bounds.

Shell inner diameter can only be between two and 4.19 inches; thus, shell inner
diameter changes to 4.19 inches. We assumed the tube cross-sectional area is a
quarter of the entire tube cross-sectional area.

It is impossible for all four passes to occupy more than one half of the entire heat
exchanger; hence, the upper bound is half the original tube cross-sectional area.

Lastly, tube pitch can be increased to an upper bound where the ratio between the
spacing and the diameters of the tubes is two.

When we change all four constraints simultaneously to the worst-case scenario, all
three Nusselt number values are halved from the values mentioned in Appendix B.2.
When we change all four constraints simultaneously to the best-case scenario, only the
tube side Nusselt number is tripled compared to the values mentioned in Appendix B.2.
Appendix B.2 is written assuming the best case scenario for the shell side and the worst
case scenario for the tube side.

5.3 Sources of Error

8
A minor source of error was fluctuations in flow rate. We used the computer program
LabView to run the heat exchanger system; LabView was not always able to keep the
flow rates constant.

Another source of error was heat transfer to the air surrounding the heat exchanger.
This heat transfer resulted in smaller temperature differences than predicted by energy
balances. From our data, we determined that approximately 3% of the heat from the
hot water flow was lost to the air.

6 CONCLUSIONS & RECOMMENDATIONS


6.1 Conclusions
As shown by our data and calculations, we determined a linear relationship between the
Nusselt number (NNu) and the overall heat transfer coefficient (U). We observed that
keeping the hot water constant and varying the cold water gave the highest U value and
a high heat transfer coefficient. Our main goal for this experiment was to learn more
about the new heat exchanger and study the how flow behaviors can be described by
Reynolds Number and Nusselt Number. The results we provide in this report are a
starting point for future experiments.

6.2 Future Lab Work


Future lab groups working on the heat exchanger should think about duplicating our
results with one variation. We suggest placing insulator around the pipes of the heat
exchanger to decrease the heat loss to the surrounding air. With insulation, there
should be a higher temperature gradient, and the overall heat transfer coefficient should
be closer to the predicted value. Further experimentation with conductive and
convective heat transfer coefficients may also improve the overall heat transfer
coefficient.

9
7 Nomenclature
Lowercase
h, convective heat transfer coefficient
k, conductive heat transfer coefficient
m, tube arrangement constant
v, average velocity
vmax, maximum velocity on the shell side

Uppercase
AT , total surface area of the heat exchanger
C, tube arrangement constant
D, tube inner diameter
NNu, Nusselt Number (tube side)
NNu,Shell, Nusselt Number (shell side)
NPr, Prandtl Number
NRe, Reynolds Number (tube side)
NRe,Shell, Reynolds Number (shell side)
Q, total heat transferred
Sn, pitch distance
T1, inlet temperature of the cold water
T’1, inlet temperature of the hot water
T2, outlet temperature of the cold water
T’2, outlet temperature of the hot water
∆ Tlm, logarithmic mean temperature
U, overall heat transfer coefficient

Greek
ρ , density
µ , viscosity
µ b, bulk viscosity
µ w, viscosity of water

10
8 References & Sources Cited
Bruce, D. Heat Exchangers. http://www.ces.clemson.edu/chemeng/uolab/heat.htm. July
26, 2000. Accessed May 14, 2001.

Bruce, D. Wilson Method. http://www.ces.clemson.edu/chemeng/uolab/wilson1.htm.


July 26, 2000. Accessed May 14, 2001.

Department of Chemical Engineering Pennsylvania State University. ChE 407W Heat


Exchanger.
http://fenske.che.psu.edu/Faculty/Matsoukas/che407/downloads/3_HE.pdf.
Spring 2001. Accessed May 14, 2001.

Geankoplis, C.J. Transport Processes and Unit Operations, 3rd Ed., PTR Prentice-Hall,
Englewood Cliffs, NJ, 1993.

Heat Exchangers. http://courses.che.umn.edu/00fscn4331-1s/heat_exchangers.htm.


Accessed May 14, 2001.

LAB 9a. Heat Exchangers. http://courses.che.umn.edu/00fscn4331-1s/lab4.htm. March


14, 2000. Accessed May 14, 2001.

McMaster-Carr. Liquid-to-Liquid Heat Exchangers (Item #35185K54).


http://www.mcmaster.com/pdf/107/0402.pdf. Accessed May 14, 2001.

Mukherjee, R. "Effectively Design Shell-and-Tube Heat Exchangers." Chemical


Engineering Progress. Vol. 94. pp. 21-37. February, 1998.

Prieve, D. Class notes, 2001.

Sandler, Stanley I. Chemical and Engineering Thermodynamics, J. Wiley and Sons,


N.Y., 1977.

University of Michigan. ChE 360 Lab I Project: Shell and Tube Heat Exchanger.
http://www.engin.umich.edu/class/che360/heatexch.html. Accessed May 14,
2001.

Wan, Wai. Heat Exchanger in Bio-Chemical Process. http://www.rpi.edu/dept/chem-


eng/Biotech-Environ/FERMENT/. Accessed May 14, 2001.

11
Appendix A: Data Manipulations
Appendix A.1: Sample Data from LabView

LabView was set up to generate a table containing sixteen values. Those sixteen
values, from left to right in the table, were:
1. Time [s]
2. Cold water flow rate [gal/min]
3. Hot water flow rate [gal/min]
4. Cold water inlet temperature for shell-and-tube heat exchanger [°F]
5. Hot water inlet temperature for shell-and-tube heat exchanger [°F]
6. Cold water outlet temperature for shell-and-tube heat exchanger [°F]
7. Hot water outlet temperature for shell-and-tube heat exchanger [°F]
8. Cold water inlet temperature for welded plate heat exchanger [°F]
9. Hot water inlet temperature for welded plate heat exchanger [°F]
10. Cold water outlet temperature for welded plate heat exchanger [°F]
11. Hot water outlet temperature for welded plate heat exchanger [°F]
12. Cold water inlet temperature for brazed plate heat exchanger [°F]
13. Hot water inlet temperature for brazed plate heat exchanger [°F]
14. Cold water outlet temperature for brazed plate heat exchanger [°F]
15. Hot water outlet temperature for brazed plate heat exchanger [°F]
16. Cold Junction (CJC) Temperature [°F]

Since we only used the shell-and-tube heat exchanger, we removed columns eight to
fifteen in the table generated by LabView. We also removed non-steady state values
via data parsing (Appendix A.2).

1
Appendix A.2: Data Parsing

Figure A.2.1 shows a graph of the data before removing non-steady-state data points.

The green oval represents data points removed for that one step change. These values
were not at steady state; during this time, LabView was changing the water flow rates.
We did not use these data points for data analysis.

12

10

8
Flow Rate [gal/min]

0
714 814 914 1014 1126 1226
Time [sec]
Cold Water Hot Water

Figure A.2.A.5: Data with Non-Steady-State Values

Figure A.2.2 shows the data in Figure A.2.1 that we used in data analysis.

2
12

10

8
Flow Rate [gal/min]

0
714 814 914 1014 1114 1214 1314
Time [sec]
Cold Water Hot Water

Figure A.2.A.6: Data Used in Data Analysis (only steady state values)

For more information on how we did the data analysis, see Appendix B.2.

3
Appendix B: Sample Calculations
Appendix B.1: Predicted Results

We used the attached MathCAD sheet, entitled “HX.mcd,” to predict the temperature
differences for various hot and cold water flow rates.

Although this MathCAD sheet predicts what to expect well, this same sheet cannot be
used to analyze the data.

1
Appendix B.2: Sample Data Computations

The attached MathCAD document, entitled “Sample_Calc.mcd,” shows the calculations


we did in order to produce the figures shown in the body of the report.

2
Appendix C: Other Significant Graphs (Experimental Data)

Figure C.7: Wilson Plot: Constant HW Flow (Tube Side)


Figure C.8: Wilson Plot: Constant CW Flow (Shell Side, square)
Figure C.9: Modified Wilson Plot: Constant CW Flow (Shell Side, Triangular)

Das könnte Ihnen auch gefallen