Professional Documents
Culture Documents
Table of Contents
Executive
Summary:
....................................................................................................................................
3
A.
Introduction
.........................................................................................................................................
4
hL major
.......................................................................................................................................................
5
hL minor
.......................................................................................................................................................
5
B.
Procedure
............................................................................................................................................
6
C.
Results ................................................................................................................................................. 7
Executive Summary:
The goal of this experiment was to determine the loss coefficients for pipe system components
and the roughness values for straight pipes using experimental data. The effects of Reynolds
number on the friction factors in straight pipes were also investigated. To perform this
experiment, a pipe system consisting of two pipe segments connected perpendicular to one
another with a 90 elbow component or a T-Bend fitting was used. Air was pumped through the
pipes at a designated flow rate and the pressure readings along the pipes were recorded using
taps connected to a digital manometer. The flow rates were determined from a set of three
Reynolds numbers to be tested. For this experiment the Reynolds numbers of 15000, 25000, and
35000 were used to determine the flow, all of which were considered turbulent. From the
pressure values obtained at these three Reynolds numbers the friction factor and roughness for
the straight pipe as well as the loss coefficient for the bend fittings were calculated. The average
friction factors for the 90 elbow and T-Bend configurations were 0.0242 and 0.0296
respectively. The 90 elbow fitting was found to have an average loss coefficient of 0.737 and
the T-Bend was found to have a coefficient of 0.873. This indicated that the 90 elbow
component is a better choice for use in a 90 angle setup as it had the lowest loss coefficient.
The resulting friction factors at different Reynolds numbers did not conclusively show how these
values change with Reynolds numbers. The values did not generally increase or decrease with
an increase in Reynolds number; according to the Moody chart, however, they should have
decreased with increasing Reynolds numbers. For the roughness, measured values were obtained
using equation D.1 and theoretical values were gathered from the Moody chart. Error
calculations showed that the results for roughness were acceptably accurate with maximum
errors of only 13% for the T-Bend. The data also showed that the higher the Reynolds number,
correspond to more accurate the roughness calculations, with an error of only 2.5% for the TBend at a Reynolds number of 35000.
A. Introduction
Turbulent pipe flow tends to be very difficult to quantitatively understand from a purely
theoretical standpoint, so experimental analysis is necessary to place some solid numerical
backing behind this theory. This lab was intended to explore the effects of various flow rates
(and consequently velocity and Reynolds number) on pressure drops through a fixed diameter
pipe system with replaceable joints. This system is illustrated in Figure 1.
hL major
The viscous effects mentioned earlier are a result of the shear stresses that exist due to interaction
between the pipe wall and the fluid flow. The pressure drop through these systems is known to
be dependent of a number of other parameters, such that:
! = !(!, !, !, !, !, !)
Here V is average velocity, D is the pipe diameter, l is the length of pipe being considered, is
the roughness of the given pipe, is the dynamic viscosity, and is the fluid density.
Two other important quantities for understanding turbulent pipe flow are the Reynolds Number,
hereafter denoted Re, and the roughness to diameter ratio, known as relative roughness, and
calculated as /D. These quantities are used in conjunction with a Moody Chart (Figure E.1) to
generate a true value for the friction factor, f, which is ultimately used to calculate the hL major.
Once the true value for the friction factor has been obtained, the calculation for hL major is:
! !"#$% = !
! !!
! 2!
hL minor
Minor losses are not necessarily minor in relation to major losses. Depending on the exact
configuration of the pipe system in question, they can actually be more significant than the major
losses. These minor losses are generated in components such as elbows, 180-degree bends, tees,
valves, and reducers. As the calculation of hL minor is difficult through traditional means, it is best
to analyze the losses through these elements by computing an equivalent straight pipe length,
denoted KL, which simulates the same reduction in flow energy as the complex element would.
This value is experimentally obtained, and is then used in the calculation of the head loss, hL
minor, through the given element. See Appendix A for detailed values of KL.
Once the correct value for KL is selected, it can be used in a formula similar to that used for
major losses, specifically:
! !"#$% = !!
!!
2!
B. Procedure
1. Initial Measurements and calculations
1.1. Read the barometric pressure prior to conducting experiment
1.2. Determine the temperature of the ambient air
1.3. Record each of the dimensions listed in Figure 1
1.4. Using calipers, measure the ID of the pipe system, and calculate the area from this value
1.5. Calculate the heights, h, on a manometer necessary to achieve a Re value of 15000,
25000, and 35000 through the pipe system.
2. Setup
2.1. Activate the manometer 1 hour prior to testing
2.2. position the yellow dump valve so that the handle is parallel to the pipe it is attached to
(this is fully open)
2.3. Press the green start button to begin running the supply fan and air conditioner
2.4. Calibrate the thermostat at the flow bench to match the temperature recorded previously
2.5. Set the multipoint selector switch set to position 1, and verify that the temperature
displayed on the thermostat still agrees with the ambient reading.
3. Calibration of Instruments and Pipe Flow rate
3.1. To zero the manometer:
3.1.1. Set the three position filter switch to OFF
3.1.2. Set the scale knob to position X1
3.1.3. Use the zeroing knob to adjust manometer until it reads 0.
3.1.4. Move three position filter switch from OFF to HI
3.2. Adjusting flow rate:
3.2.1. position the yellow dump valve so that the handle is parallel to the pipe it is
attached to (this is fully open)
3.2.2. Verify that the valve is open by placing a hand in front of the opening following
the valve, and noting the presence of airflow.
3.2.3. Use the blue dial valve to adjust the flow within the system so that the height, h,
on the manometer reads the appropriate value calculated in step 1.5. Note that if the
correct reading cannot be achieved, slowly close the yellow dump valve until the
reading on the manometer is above the desired reading, then bring back down using
the blue dial valve
4. Pipe A: 90 Degree Elbow Fitting
4.1. Adjust flow until the manometer reads the height that corresponds to Re = 15000, using
the techniques listed in steps 3.2.1 -3.2.3
4.2. Set the knob labeled HP, the high pressure source, to PIPE A
4.3. Set the knob labeled LP , the low pressure source, to R1
4.4. Verify that the position of the handle on the ball valve for pipe A is parallel to the pipe
(fully open), and that the handle on pipe B is perpendicular to the pipe (fully closed)
4.5. Set the selector knob for Pipe A to position 1
4.6. Record the differential pressure displayed on the electronic manometer.
4.7. Repeat steps 4.5 and 4.6 for ports 2-11
5. Pipe A: 90 Degree Tee Fitting
5.1. Replace the 90 degree fitting the with the capped Tee fitting. Verify that the joints are
properly sealed by feeling for any airflow around the joints
5.2. Repeat steps 4.5 -4.7 for this fitting.
6. Shut Down
6.1. Press the red stop button for the fan and air conditioner unit
6.2. Close the blue dial valve completely
C. Results
The raw data and Reynolds number calculations are listed in Appendix A. The pressure at each
tap along with the geometry of the test setup was used to analyze the flow of air through the pipe
and 90 degree elbow and T bends.
C.1. Pipe A: 90 Degree Elbow Fitting
The normalized raw data found in Appendix B was used to create a plot of pressure versus length
along pipe seen in Figure 2.
Figure 2. Plot of Pressure versus length along pipe for 90-degree elbow setup.
7
The friction factor for each Reynolds number and average friction factor are listed in Table 1.
Details of the calculations of the friction factor are contained in Appendix C.
Table 1. Friction factors for the 90 degree elbow bend setup
Reynolds Number
Friction Factor,
15000
25000
35000
Average
0.0239
0.0266
0.0221
0.0242
Table 2 contains values for the theoretical and measured relative roughness and roughness for
each Reynolds number and the percent error between the theoretical and measured roughness
values. Appendix D contains details of the roughness calculations.
Table 2. Theoretical and measured values of /D and for the 90 degree elbow bend
Reynolds Estimated /D
Number
15000
25000
35000
0 (Smooth)
0.00080
0 (Smooth)
Calculated /D
Theoretical
(mm)
Measured
(mm)
Percent
Error (%)
-0.0185
0.00090
-0.00019
0.0
0.0212
0.0
-0.0490
0.0238
-0.00500
--12.3
---
The calculated values of the loss coefficient for the 90 degree elbow bend are contained in
Table 3. Details of these calculations can be seen in Appendix E.
Table 3. Loss coefficient values for the 90 degree elbow bend
Reynolds
Number
Measured Loss
Coefficient, KL
15000
25000
35000
Average
0.500
0.839
0.873
0.737
Figure 3. Plot of Pressure versus length along pipe for T bend setup.
The friction factor for each Reynolds number and average friction factor are listed in Table 4.
Details of the calculations of the friction factor are contained in Appendix C.
Table 4. Friction factors for the T bend setup
Reynolds Number
15000
25000
35000
Average
Friction Factor,
0.0325
0.0266
0.0296
0.0296
Table 5 contains values for the theoretical and measured relative roughness and roughness for
each Reynolds number and the percent error between the theoretical and measured roughness
values. Appendix D contains details of the roughness calculations.
Table 5. Theoretical and measured values of /D and for the T bend setup
Reynolds Estimated /D
Number
15000
25000
35000
0.0025
0.00080
0.0030
Calculated /D
Theoretical
(mm)
Measured
(mm)
Percent
Error (%)
0.00281
0.00089
0.00308
0.0663
0.0212
0.0795
0.0745
0.0237
0.0815
12.9
11.8
2.52
The calculated values of the loss coefficient for the T bend are contained in Table 6. Details
of these calculations can be seen in Appendix E.
Table 6. Loss coefficient values for the T bend
Reynolds
Number
Measured Loss
Coefficient, KL
15000
25000
35000
Average
0.707
1.09
0.835
0.878
Friction Factor,
Roughness, (mm)
Loss Coefficient
Pipe
T-Bend
90 Degree Elbow
0.0269
-----
0.0 0.0815
-----
--0.878
0.737
D. Discussion
The goals of this lab were to determine the effects of Reynolds number on the friction factor for
straight pipes and to find the loss coefficients for both a 90 elbow and T-bend fitting. The
roughness for the straight pipe was also calculated. The calculated results for the roughness were
fairly accurate with the results obtained from the literatures Moody Chart, with errors less than
13%. The results were inconclusive as to what effect increasing the Reynolds number had on the
friction factor. According to the Moody chart, the friction factor should decrease with increasing
Reynolds numbers; however, the calculated friction values from the experiment increased from
15000 to 25000 and then decreased from 25000 to 35000. This discrepancy from the behavior
predicted in the literature was most likely due to the many sources of error present in the
experiment.
One possible source of error came from the pressure readings. The pressure readings at the
reference point for each component and each flow was some value greater than zero (Tables A.5
and A.6), but the problem was that all the reference point readings should have been zero
regardless of the set up. The reason for this discrepancy remains undetermined, however it is
suspected that there was a problem with the machines manometer. This theory was supported
by the fact that when taking readings at the various reference taps, the manometer value never
10
stabilized; instead, it would often slowly decrease, increase, or bounce around randomly.
Readings were taken when the pressure value shown appeared the most stable, however no
reading was ever truly stable. Despite this problem, the data obtained was still used, and to
account for the reference pressure not being zero, the data was normalized (Table B.1 and B.2).
The P values obtained from the data should still be the same, with or without normalizing, so
the high reference pressure should not affect the friction factors calculated. The real problem
with having incorrect reference pressures was the aforementioned implication that the
manometer took incorrect readings. That combined with the unstable pressure readings is where
the error would come from. It is hard to say precisely in what matter or how significantly these
incorrect readings affected the results and the values calculated from the results since the
pressure readings were not consistently incorrect. For instance, when the pressure readings
constantly decreased at one reference tap, then the overall P was probably higher than it should
have been for that section, but if the pressure readings were constantly increasing, then the P
would be smaller than it should have been, making the friction factors higher and lower for each
case respectively. It is likely that this problem caused the bulk of the error of the data, perhaps
causing as much as 50% of the error for the lab.
Another source of error came from the connections between the two different branch
components, the T-bend and the 90 elbow. When components were changed it was necessary to
attempt to perfectly line up both Pipe A and the pipe segment after the branch component with
the openings of the branch component. However, these connections did not line up perfectly
every time, which may have caused small air leakages at the connections. These small leaks
would lead to a greater pressure drop across the component than the ideal situation. According to
equation F1, KL is proportional to P, which means that the increased P across the elbow or Tbend would have resulted in a KL value that was higher than expected. However, the connections
were inspected before each test to verify that no air could be felt escaping from the pipes. This
ensured that any leaks from the connection points were very small, so it is possible that this error
source did not contribute to more than 30% of the errors in our KL values.
The misaligned branch component would also have caused an increase in P because parts of the
pipe ends would be jutting into the airstream at the connection points. These protrusions into the
airstream would have obstructed some of the flow, leading to an increased drop in pressure
across the branch component. As stated above, this greater value for P would have resulted in a
higher experimental KL value. It is difficult to ascertain whether this source of error or the
leakages contributed more to the higher-pressure drops measured, though due to the fact that no
obvious leaks were observed it can be predicted that a greater amount of the error in the KL
results was due to the misalignment of pipes.
11
Length (mm)
L1
L2
L3
L4
L5
L6
LU
LD
Tap Spacing
25
840
940
840
25
940
253
251
10
26.5
Table A.2 contains the measurements of atmospheric pressure and temperature at the time of the
experiment.
12
Atmospheric
Pressure (kPa)
Ambient
Temperature (F)
Ambient
Temperature (K)
758.4
101.1
75.2
297
Data was to be taken for flow with Reynolds values of 15,000, 25,000, and 35,000. In order to
determine the flow rates necessary to produce these values, the cross-sectional area of the pipe
and the necessary air velocity was calculated by Equations A.1 and A.2 respectively.
A.1
A.2
In these equations D is the pipe diameter, Re is the desired Reynolds number, and is the
kinematic viscosity of air determined by Equation A.3, known as the Sutherland Equation, and
Equation A.4.
A.3
A.4
In these equations is the dynamic viscosity of air, T is the temperature in Kelvin, is the
density of air, and C and S are constants. C and S were determined from known values of at
certain temperatures to be:
The density of the air was determined by the ideal gas equation in Equation A.5.
A.5
In this equation P is the atmospheric pressure in Pascals, T is the temperature in Kelvin, and R is
the gas constant of air. Table A.3 contains the value of R along with the density and kinematic
viscosity of air at the temperature and pressure the experiment was performed at.
Table A.3. Gas constant and Kinematic and dynamic viscosities of air
R (J/(kg*K))
Density, (kg/m3)
287.05
1.54*10-5
1.186
13
Once the values of velocity and area were determined, the flow rate, Q, was found from Equation
A.6.
A.6
The required flow meter pressure, measured in millimeters of mercury, was then determined
from the relationship in Equation A.7.
A.7
The values for velocity, flow rate, and flow meter pressure required for each Reynolds number
are contained in Table A.4.
Table A.4. Velocities, flow rates, and pressure for each Reynolds number
Reynolds Number
Velocity (m/s)
Q (m3/s)
H (mm Hg)
15000
25000
35000
9.21
15.19
21.17
0.0051
0.0084
0.0117
20.0
33.0
46.0
Tables A.5 and A.6 contain the raw data as recorded in the lab for the flow with each Reynolds
number for the 90 degree elbow bend and T bend setup respectively.
Table A.5. Raw data for the 90 degree elbow bend setup
Reynolds Number
Flow Meter, H (mm H2O)
15000
25000
35000
20
33
46
Tap Number
1
2
3
4
5
6
7
8
9
10
11
1.10
1.05
1.00
0.96
0.92
0.92
0.56
0.52
0.49
0.46
0.44
2.66
2.55
2.45
2.33
2.25
2.15
0.77
0.65
0.54
0.45
0.35
14
2.66
2.55
2.45
2.33
2.25
2.15
0.77
0.65
0.54
0.45
0.35
15000
25000
35000
20
33
46
Tap Number
1
2
3
4
5
6
7
8
9
10
11
3
2.9
2.81
2.71
2.6
2.5
0.86
0.74
0.63
0.54
0.43
15
5.58
5.35
5.15
4.9
4.6
4.39
1.6
1.35
1.14
0.98
0.76
15000
25000
35000
0.0051
0.0084
0.0117
Distance (m)
1
2
3
4
5
6
7
8
9
10
11
0.00
0.100
0.200
0.300
0.400
0.500
1.004
1.104
1.204
1.304
1.404
308.0
293.3
280.0
264.0
253.3
240.0
56.0
40.0
25.3
13.3
0.0
16
544.0
521.3
502.6
481.3
460.0
440.0
96.0
69.3
42.7
25.3
0.0
15000
25000
35000
0.0051
0.0084
0.0117
Distance (m)
1
2
3
4
5
6
7
8
9
10
11
0.00
0.100
0.200
0.300
0.400
0.500
1.004
1.104
1.204
1.304
1.404
308.0
293.3
280.0
264.0
253.3
240.0
56.0
40.0
25.3
13.3
0.0
17
544.0
521.3
502.6
481.3
460.0
440.0
96.0
69.3
42.7
25.3
0.0
Figure C.1.Plots of pressure versus length for the 90 degree elbow bend setup.
18
Figure C.2. Plots of pressure versus length for the 90 degree elbow bend setup.
In this experiment the pipe flow between taps other than 6 and 7 (where the bend was located)
was modeled by Equation C.2.
C.2
The slope of an equation of a line of best fit corresponds to a value of
since the plots are of
pressure versus length. Values of the friction factor, , were found by taking the average of the
slopes from each segment (taps 1 through 6 and taps 1 through 7) for each Reynolds number and
then dividing by (-V2)/(2D), where , V, and D are all known values. The average of theses
values was then taken to find the average friction factor. Table C.1 and C.2 contain values for
the slope of each line segment, the average slopes, the values of (V2)/(2D), and the friction
factors for each Reynolds number for each setup as well as the average friction factor.
19
Slope of
taps 1-6
15000
25000
35000
-50.7
-136.0
-207.2
Slope of
taps 7-11
-40.0
-138.7
-236.0
Average
slope, mavg
(-V2)/(2D)
(N/m3)
-45.3
-137.3
-221.6
-1897
-5165
-10036
0.0239
0.0266
0.0221
Average
0.0242
Slope of
taps 1-6
Slope of
taps 7-11
Average
slope
(V2)/(2D)
(N/m3)
15000
25000
35000
-70.1
-133.3
-321.8
-53.3
-141.3
-273.3
-61.7
-137.3
-297.6
-1897
-5165
10036
0.0325
0.0266
0.0296
Average
0.0296
The overall average friction factor was found by averaging the average friction value from each
setup. This value was determined to be:
= 0.0269
20
Table D.1. Measured and Theoretical values of /D and for 90 degree elbow bend
Reynolds
Number
15000
25000
35000
Theoretical
Measured
Relative
Relative
Roughness, /D Roughness, /D
0 (Smooth)
0.00080
0 (Smooth)
Theoretical
Roughness,
(mm)
Measured
Roughness,
(mm)
Percent
Error
(%)
0.0
0.0212
0.0
-0.0490
0.0238
-0.00500
--12.3
---
-0.00185
0.000897
-0.000189
Theoretical
Measured
Relative
Relative
Roughness, /D Roughness, /D
0.0025
0.00080
0.0030
Theoretical
Roughness,
(mm)
Measured
Roughness,
(mm)
Percent
Error
(%)
0.0663
0.0212
0.0795
0.0745
0.0237
0.0815
12.9
11.8
2.52
0.00281
0.00089
0.00308
21
22
Reynolds
Number
PBefore (Pa)
PAfter (Pa)
P (Pa)
2/(V2)
KL
15000
25000
35000
52.5
205.2
383.9
27.4
90.5
151.6
25.1
114.8
232.3
0.0199
0.0073
0.0038
0.50
0.84
0.87
Average KL
0.74
Table F.2. Pressure drop and Loss Coefficient values, KL, for T bend
Reynolds
Number
PBefore (Pa)
PAfter (Pa)
P (Pa)
2/(V2)
KL
15000
25000
35000
72.4
241.2
408.7
36.8
91.8
186.7
35.6
149.4
222.0
0.0199
0.0073
0.0038
0.71
1.09
0.83
Average KL
0.88
23
Appendix G: References
Munson, Bruce Roy, Donald F. Young, and T. H. Okiishi. Fundamentals of Fluid
Mechanics. Hoboken, NJ: J. Wiley & Sons, 2009.
24
25