DEPARTMENT OF CIVIL ENVIRONMENTAL & GEOMATIC ENGINEERING
FLUID MECHANICS II
PRESSURE AROUND A CYLINDER AND CYLINDER DRAG
OBJECTIVES
1. To measure the pressure around the circumference of a circular cylinder in a uniform
steady air flow for different flow velocities and to plot diagrams for distribution of the
corresponding pressure coefficients.
2. To compare the measured pressure coefficient distribution with standard measurements in
the literature (Figure 3).
3. To integrate the distribution of measured pressure coefficient around the cylinder and
obtain the form drag coefficient. To plot the relation between the drag coefficient and the
Reynolds number and to compare this with classical results (Figure 2).
4. To measure the velocity profile in the wake downstream of the cylinder. To understand
the qualitative relation between the form of this profile and the drag.
5. To use spreadsheets for collecting, processing and analysing experimental data.
Lab reports must be prepared individually and submitted before the specified deadline.
THEORY AND DEFINITIONS
1. Drag force
The force exerted on a body by a flowing fluid and acting in the direction of flow is called drag.
The drag is due to pressure and shear stresses applied by the fluid to the surface of the body.
The part of the drag associated with pressure very much depends on the shape of the body and
is often called form drag. For bluff bodies, form drag can constitute a considerable part of
total drag. The form drag can be obtained by integration of pressure acting on a projection of
the body perpendicular to the flow direction over all the surface of the body. For a cylinder
perpendicular to the flow (Figure 1) the elementary drag force per unit width at angle α is
dD = P cos α r dα
P dl = r d α
dα
U
D α O A
r
Figure 1
1
and the total drag per unit width can be obtained by integration over the cylinder circumference.
The flow is symmetric and pressure distributions on the upper and lower surfaces are the same.
Thus, we can integrate from 0 to π and multiply the result by two. We have:
Zπ
D = 2r (P − P∞ ) cos α dα . (1)
0
Here we subtracted the pressure in the flow far from the body, which is often used as the reference
pressure.
2. Drag coefficient and pressure coefficient
It is convenient for both practical applications and theoretical analysis to describe physical
phenomena in non-dimensional form. This allows us to find universal relations which can be
applied to different physical conditions, for example to experimental models and real life proto-
types. Such a non-dimensional parameter describing drag of bodies in a fluid flow is the drag
coefficient. Using your common sense and everyday experience you could deduce that drag
should depend on flow velocity (higher velocity → higher drag) and shape and size of a body
(larger body → higher drag). The drag force can be represented as the product of some averaged
pressure by the area of body projection perpendicular to the flow. For example, for a cylinder
perpendicular to flow, this area is A = d b, where d is cylinder diameter and b is span. When
the cylinder is very long the flow can be considered two-dimensional. In this case we can use
drag per unit width and take b = 1 unit. For identical bodies the pressure distribution depends
on fluid velocity. The value of ρ U∞ 2 has the swame dimension as the pressure, and one can
2
suggest that P ∼ ρ U∞ . We can see now that the drag can be represented as D ∼ ρ U∞ 2 A,
and the corresponding non-dimensional coefficient of proportionality is the drag coefficient. The
conventional definition for the drag coefficient is
D
CD = 2 /2
, (2)
A ρ U∞
where U∞ is the velocity of the uniform flow far from the body, ρ is the fluid density and A is
the area of a body projection perpendicular to velocity.
For geometrically similar bodies in the flow of an incompressible viscous fluid the drag
coefficient depends only on the Reynolds number, which describes the importance of viscosity.
For our case the Reynolds number can be specified as
ρ d U∞
Re = ,
µ
where ρ and µ are the density and dynamic viscosity of air, d is cylinder diameter and U∞ is the
free stream velocity. Therefore, specifying the particular form of the relation of drag coefficient
with Reynolds number CD = CD (Re) is a problem of great practical importance. By using a
single plot of CD as a function of Re one can calculate drag for similar bodies of various sizes
and for various flow velocities from the formula
ρ U∞2
D = CD A. (3)
2
Wind tunnel experiments is one of the conventional methods of obtaining the relation of CD
from Re. Examples of experimental results for the relation of pressure coefficients to Reynolds
number for different body shapes are represented on Figure 2.
2
3
Figure 3 b
Figure 2a
a
From Massey,B.S., ”Mechanics of Fluids”
b
From Nakayama,Y. Boucher,R.F. ”Introduction to Fluid Mechanics”
Similarly, pressure distribution over a body surface can be specified via the non-dimensional
pressure coefficient
P − P∞
CP = 2 /2
. (4)
ρ U∞
Pressure coefficients around similar bodies of different sizes in flows of different velocities will
be the same if the Reynolds numbers of these flows are the same, although the corresponding
pressures can be different. An example of pressure coefficients around a cylinder for different
Reynolds numbers is shown in Figure 3
To measure pressure coefficients in experiments one should measure pressure difference be-
tween two static pressure tapings, one being on the surface of the cylinder and the other in the
area of undisturbed flow, ideally far upstream of the cylinder.
It is convenient to use the measurement at the front stagnation point (point O, Figure 1),
where velocity is zero, to calculate the velocity of the undisturbed flow. The Bernoulli equation
along the streamline AO reads:
ρ U∞2
P∞ + = P0 .
2
This gives
ρ U∞2
= P0 − P∞ ,
2
which is simply the manometer reading at α = 0. Then to obtain the pressure coefficient at any
α it is sufficient to divide the corresponding manometer reading by the stagnation point reading
P (α) − P∞
CP (α) = .
P 0 − P∞
The stagnation point reading can also be used to estimate the undisturbed flow velocity:
p
U∞ = 2 (P0 − P∞ )/ρ .
Combining equation (1) with (4) and (2) you can find that the drag coefficient can be obtained
by integrating the pressure coefficient over the cylinder surface:
Zπ
CD = CP (α) cos α dα . (5)
0
3. Using a Pitot-static tube for measuring flow velocity
To manometer
A Pitot-static tube is the combination of a
P P0 static pressure tapping with a total pres-
sure tapping (Figure 4). According to the
Bernoulli equation the pressure at the stag-
nation point (the total pressure) is P0 =
P + ρ U 2 /2, and measuring the difference
U P
Static pressure
between P and P0 one can calculate the
P0 Total pressure velocity as
p
U = 2 (P0 − P )/ρ
Figure 4
4
C
B A
Figure 5 Figure 6
C
A
B
B
C
U D
Figure 7 Figure 8
B C D
A
E F G
H I
J
Figure 9 Figure 10
5
A
Figure 11 Figure 12
EQUIPMENT
Experiments are performed in the wind tunnel of the Civil Engineering Fluids Laboratory (Fig-
ure 5). A cylinder, 114mm or 60mm in diameter, is placed across the working section (Fig-
ure 7,8). A fan (A) powered by an electric motor creates an air flow along the tunnel. Flow
velocity U∞ can be regulated by changing fan shaft power from 20% to 100%. Static pressure
tappings on the cylinder surface (C) and on the floor of the working section (D) are used to
measure the difference between pressures on the cylinder’s surface and in the uniform flow. A
Pitot tube (B) is used to measure the flow velocity in the wake downstream of the cylinder.
Pressure differences are measured by a micro manometer (Figure 6 A). The analogue voltage
signal proportional to pressure difference is transmitted from the micro manometer (A) to the
digital multimeter (B) and then to a computer (C) to be acquired by acquisition software. The
data then can be copied to a spreadsheet for processing and analysis.
EXPERIMENTAL PROCEDURE
1. Before you start
1. Read the Fluids Laboratory safety rules. Read and sign the Risk Assessment Form. Make
sure that you understand possible risks and risk control measures.
2. - Set the pneumatic valve switch of the micromanometer (Figure 9B) to ”= ” (up
position).
- Set the operation mode switch of the micromanometer (C) to ”±∆P ” (down posi-
tion).
- Set the pressure range switch (D) to the desired position. Use ”1%” for fan power
less than 60% and ”10%” othervise.
- Set the automatic zero switch to ”AUTO ” (up position).
6
- Check that plastic pressure tubes are connected to the positive and negative air
pressure ports (J). Make sure that the high pressure tube (red) is connected to the
positive (right) port and the low pressure tube (yellow) is connected to the negative
(left) port.
- Turn on the mains switch (F), the green indicator light should glow.
- Use the ”ZERO ” control (G) to set zero on the micromanometer indicator (A). It
helps to have the ”TIME CONSTANT ” control (H) in the fully anti-clockwise position
at this stage.
- Set the ”TIME CONSTANT ” control (H) to the left horizontal position.
WARNING! THE PNEUMATIC VALVE SWITCH (B) MUST BE SET TO ”=” (UP
POSITION) ALL THE TIME WHEN READINGS ARE NOT TAKEN. DO NOT
SQUEEZE OR APPLY PRESSURE TO THE PLASTIC PRESSURE TUBES.
3. Switch on the multimeter, check that DC and V options are chosen. Use ”∨ ” and ”∧ ”
buttons to set 3 digits after the decimal point on the multimeter digital indicator. Note
that the voltage output of the micromanometer is ±5V for the full pressure scale, which
is 100, 10 or 1 mm of water depending on the position of the pressure range switch of the
micromanometer. Check that the error in the multimeter reading is within 2% of the full
voltage range.
4. Switch on the computer. Start the “PC1604” data acquisition program (Figure 10). Set
“Log interval” to 2 seconds and “Log Count” to 9. Press “START” and then “Cancel”
on the file opening prompt. Nine subsequent readings of the multimeter with 2 seconds
interval will be acquired and appear in the ”LOGGER ” window of the acquisition program.
5. Open the “Cylinder Data” spreadsheet in the folder C:\Cylinder Experiment and choose
the worksheet according to your fan power. The spreadsheet is developed for collecting
and primary processing of your experimental data. Study it carefully. Average manometer
readings and the corresponding pressures will be calculated automatically after entering the
data from the acquisition program into the spreadsheet. Check that formulas calculating
the values of average readings, pressures and the Reynolds number are correct. To obtain
the values for velocities and pressure coefficients, and to get the points on the graphs you
should insert correct formulas into the corresponding rows of the table.
Note: This spreadsheet is an example how Excel can be used for processing and analysing
experimental data. You should be able to develop similar spreadsheets.
Try copying data from the window of the acquisition program to the spreadsheet. See
Appendix 1 for details.
6. Use the optical level (Figure 11) to set the position of the Pitot-static tube at the level
with the cylinder axis. Fix the slider of the Pitot tube ruler (Figure 12 A) at 40cm.
7. Switch on the wind tunnel motor and gradually increase the power to the required level.
The flow velocity at the position of the cylinder for each power can be estimated by the
following calibration relation
U = 0.079 P − 0.9 , (6)
where U is the velocity of the flow in m/s and P is the fan power in %.
7
2. Pressure around the cylinder
1. Check that the pressure tubes from the cylinder tapping and from the steady flow tapping
are connected to the micromanometer.
WARNING! THE PNEUMATIC VALVE SWITCH (Figure 9B) MUST BE SET TO
”=” (UP POSITION) WHEN PERFORMING ANY OPERATIONS WITH PRES-
SURE TUBES.
2. Set the cylinder pressure tapping at the upstream stagnation point (α = 0◦ ).
3. Set the pneumatic valve switch on the micromanometer panel (Figure 9B) to ”READ ”
(down position).
Note: Wait at least 20 seconds before starting data acquisition.
4. Click the “START” button on the acquisition program window (Figure 10) and then
“Cancel” on the file opening prompt. Wait until readings are acquired.
5. Close the pneumatic valve switch (up position).
6. Transfer data to the spreadsheet (see Appendix 1).
7. Rotate the cylinder anti-clockwise with 15 degrees interval and repeat measurements for
points over the upper cylinder surface including the rear stagnation point (α = 180◦ )
3. Velocity profile in the wake
1. Connect the pressure tubes from the Pitot-static tube to the micromanometer. Make sure
the the pneumatic valve switch (Figure 9B) is closed (up position).
2. Check that the Pitot-static tube is on the level with the cylinder axis and the reading of
the ruler is 40cm (Figure 12).
3. Open the pneumatic valve switch on the micromanometer panel (Figure 9B).
Note: Wait at least 20 seconds before starting data acquisition.
4. Click the “START” button on the acquisition program window (Figure 10) and then
“Cancel” on the file opening prompt. Wait until readings are acquired.
5. Close the pneumatic valve on the micromanometer panel.
6. Transfer data to the spreadsheet (see Appendix 1).
7. Release the screw of the Pitot-tube holder (Figure 12B). Carefully move the Pitot-static
tube and fix it at new position according to entries of the spreadsheet table. Repeat
measurements for all coordinates shown on the table.
4. After you finish
1. Switch off the wind tunnel fan.
2. Assess your experimental data.
- Check that all data in the header of your spreadsheet are correct.
- Check that the measured uniform flow velocity is reasonably close to a value obtained
by the calibration relation (6).
8
- Check that the Reynolds number is calculated correctly.
- Insert appropriate formulas for pressure coefficients and velocity into the correspond-
ing rows of the spreadsheet. Check if your results appear in the ”Results” worksheet.
- Compare the plot for pressure coefficient with the standard result on Figure 3.
- Discuss with your colleagues the shape of the velocity profile in the wake. Does it
look reasonable?
3. Switch off the equipment. Make sure that the pneumatic valve switch of the micromanome-
ter is closed (up position).
Note: Your results and results of other groups will be available on the course web page for
further analysis and preparation of lab reports.
REPORT WRITING AND DISCUSSION
1. Use the spreadsheet with your own experimental results and the two spreadsheets sum-
marising all groups results for preparing your report. Study the summary spreadsheet and
find out how it was created. If you feel confident, you can develop a similar spreadsheet
of you own.
2. Include the graph of Cp versus α to your report. Compare this graph with the standard
results on Figure 3. Comment on the shape of the graph and on differences between you
results and standard results.
3. Using the data from the summary spreadsheets calculate the drag coefficient using equation
(5) and the drag per 1m span of the cylinder by equation (3) for two cylinder diameters.
A method of numerical integration using spreadsheets is discussed in Appendix 2. You
can use the template provided in the “Integration template” worksheet.
4. Plot (i) drag D as a function of free-stream velocity U∞ for two cylinder diameters on
one plot and (ii) drag coefficient CD as a function of Reynolds number for two cylinder
diameters on another plot. Compare the graphs and discuss which of them is more useful
for practical applications.
5. Compare the observed values of the drag coefficient with the standard results presented
on Figure 2. Comment on differences between the measured and standard results. Is the
drag due to shear stresses important for the studied range of Reynolds numbers? Discuss
the variations of measured drag coefficients with Reynolds number.
6. Include the graph of the wake velocity profile in your report. Comment on the shape of
the graph. Why is the velocity in the wake lower then the free-stream velocity? Why is
the uniform velocity outside the wake higher than the free-stream velocity?
7. Give a general discussion of your results. Discuss the physics of drag generation. Why
is the representation of drag in the non-dimensional form of drag coefficient important?
What is the role of experiments in obtaining universal relations of drag coefficients to
Reynolds number? What is the practical importance of such relations?
8. Apply the results of your experiment to the following practical problem:
A radio mast has a 5m long cylindrical section 0.2m in diameter. Find the wind speed
when the results of your experiment can be used for calculating wind load on the mast.
Calculate the total load on the section at this speed assuming a uniform wind profile.
9. Make general conclusions to your report.
9
APPENDIX 1
Copying data from the acquisition program to the spreadsheet
1. Select logger output on the panel of the “PC1604” acquisition program and copy it to the
buffer (Ctrl+c).
2. Select appropriate cells of the spreadsheet and paste the data from the buffer (Ctrl+p).
Select “Use text import wizard” in “Paste Options”.
10
3. Select “Fixed width”. Press “Next”
4. Set column breaks to include numerical data (including minus signs) into one column.
Press “Next”.
11
5. Select “Do not import column (skip)” for all columns not containing numerical data. Press
“Finish”.
6. Numerical data are now imported to the spreadsheet. Average reading and the pressure
difference will be calculated in the corresponding cells.
7. Save the changes (Ctrl+s).
12
F(x) F(x4 ) F(x5 )
F(x3 ) A F(x6 )
A4
F(x2 ) A3 A5
F(x1 ) F(x7 )
A2 A6
A1
x1 x2 x3 x4 x5 x6 x7 x
a b
Figure 13
APPENDIX 2
Using spreadsheets for numerical evaluation of integrals
Rb
A value of a definite integral I = a F (x) dx of some function F (x) from x = a to x = b
can be defined as the area between the graph of this function and an x-axis on the interval
a ≤ x ≤ b. If the value of F (x) is known at n points x1 , x2 , x3 , . . . xn on the interval a ≤ x ≤ b
including points a and b, the value of the integral can be approximately found as sum of areas of
trapezoids A1 , A2 , . . . An−1 , as shown on Figure 13 for the case of n = 7. Representation by a
sum is convenient for numerical evaluation of integrals by computer, and the method described
here is called the trapezoidal rule. The evaluation formula for the trapezoidal rule is:
n−1
X 1
I≈ ( F (xi+1 ) + F (xi ) ) ( xi+1 − xi ) . (7)
2
i=1
The accuracy of approximation increases for larger number of points n.
An example below shows how Excel spreadsheets can be used for evaluation of integrals. We will
√
evaluate integrals of functions F (x) = x and G(x) = x2 from 0 to 10 using 11 points evenly
distributed over the interval. First table on the spreadsheet includes values of x in cells B2-L2,
and values of functions F (x) and G(x) at these points in cells B3-L3 and B4-L4 respectively:
Cells C6-L6 and C7-L7 include values of individual additives in the sum (7) and zeros are inserted
to cells B6 and B7 to preserve the length of the rows. Formulas C6=(C2-B2)*(B3+C3)/2 and
C7=(C2-B2)*(B4+C4)/2 are inserted manually and then copied to the rest of the row:
13
After this sums of values in B6-L6 and B7-L7 are calculated and written to the cells M6 and M7
respectively, which are the required numerical estimates of integrals:
The number of points in the example is small, and the approximation is rather rude. However
the obtained estimates 20.89 and 335.00 are fairly close to the exact values 21.08 and 333.33.
Better approximations can be obtained by taking larger number of points. A different example
and an integration template are provided in the lab spreadsheet.
14
READING
Massey,B.S. Mechanics of Fluids, 8th edition, Taylor & Francis, 2006.
8.8.3 Components of drag
8.8.4 Profile drag of two-dimensional bodies
3.7.1 The Pitot tube and the Pitot-static tube
15
