Water Tunnel Experiment MAE 171A/175A/126

Laboratory Manual

Part 1: Water Tunnel Velocity Calibration and Drag Measurement

on a Cylinder From Pressure Measurements

Objectives:

a)Calibrate the velocity of the water tunnel as a function of impeller motor rpm for
use in subsequent hydrodynamic experiments using static pressure from pitot
tubes
b)Make measurements of the drag coefficient of a circular cylinder at three different
flow velocities (0.6, 1.0 & 1.4 m/s). Both the velocity calibration and
drag measurements are based on pressure measurements.

Discussion:

A circular cylinder is located vertically in a water tunnel. A pressure sampling system

provides measurements of the pressure different across a pitot static probe and the
pressure distribution around the 1-inch diameter (2.54 cm) cylinder.

The drag Dp due to the pressure distribution on the cylinder can be calculated by
integrating the pressure p around the cylinder:

Dp = ∫ p × cosθ × da (1)

The pressure Drag can be rewritten in terms of the pressure drag coefficient CDP,

€
CDp =
Dp
=
∫ pcosθda = ∫ C p cosθda
1 1 A
ρU 2 A ρU 2 A
2 2 (2)

Where A is the reference area and Cp is the pressure coefficient,

€ p − pref
Cp =
1
ρU 2
2 (3)

To perform the integration in eq.(1) around the cylinder we integrate in terms of the
angular coordinate θ from 0 to 2Π. The angular coordinate θ, is taken to be zero at
€ the leading edge stagnation point. The reference area is A= D * b where D is the
diameter of the cylinder and b is a unit length along the cylinder, giving the
incremental area da = (D/2) * b dθ. The pressure drag coefficient can then be
calculated from the following integral.
 1 2Π ⎛ D ⎞ 2Π
CDp = ∫ C cosθb⎜⎝ 2 ⎟d
bD 0 p ⎠
θ = ∫ C p cos θdθ
0 (4)

Experimental Setup:
€
The Water Tunnel is 6 Feet long. It has a square cross section with width being 10
MAE 17 IA1175A112h Water Tunnel E a p d e n t W~nter2W6
Inches. The pressure distribution system consists
R.J. Canolica, ofG.R.
modified by a Tynan
1” cylindrical test specimen that
has been configured with 13 pressure ports. A pitot tube at the inlet of the tunnel
monitors the dynamic pressure of the flow. A computer that has an analog to digital
converter, digital input and output ports and Lab-view software installed controls the
system. The 13 cylinder pressure ports and the pitot stagnation port are routed to 14
solid-state pressure transducers that convert the mechanical pressure to an analog
electrical signal. A multiplexer selects one port under computer control and routes the
pressure value to the computer’s A to D converter input. The signal is then converted by
Experimental Procedure:
the analog to digital converter into a digital value. All the ports are scanned and the
1 .Calibrate the pressures are
water tunnel test shown
section on a manometer
by generating a plot of velocitytype
versusdisplay to visualize pressure distribution
motor frequency
using the upstream
around 180 degrees
pitot-static of the cylinder.
tube and Bernoulli's equation. The portthe
Calibrate pressure of the
operation samples are then recorded to disk
motor frequencies from 15 to 45 Hz in 5 Hz increments. A calibration of the
water tunnel forwhen the operator selects record data for post processing.
pressure transducer will be provided in the laboratory.

1 1
Pa + ρaU a2 = Pb + ρbU b2
2 2
Ua = 0 & Ub = U∞

1
Pa − Pb = ρ∞U ∞2 = q∞
2
∴Pa − Pb = ρ1gh

1
⎛ 2( P − P ) ⎞ 2
a b
P1
U ∞ = ⎜⎜ ⎟⎟
⎝ ρ ⎠
Pitot Probe: Measorempnt oCFlnid Vdoclty
1
Pitot Tube: Measurement of Fluid Velocity ⎛ 2 ρ Δhg ⎞ 2
∴U ∞ = ⎜ 1 ⎟
⎝ ρ∞ ⎠
where:
Experimental Procedure:
€ generating a plot of velocity versus
1. Calibrate the water tunnel test section by
motor frequency using the upstream pitot-static tube and Bernoulli's equation.
Calibrate the operation of the water tunnel for motor frequencies from 5 to 45
and:
Hz in 5 Hz increments.
2. Turn the power to the computer on and allow the system to boot into
windows.
3. Turn on the power to the Toshiba power inverter for the motor.
4. Ensure the power indicator is lit on the aluminum box with the red neon lamp
(located on the lower shelf beneath the bench).
5. Follow the instructions below to run the calibration and cylinder pressure
distribution measurement experiment.
a. Calibration
I. In the PressureDistribution folder on the desktop, double click on
WatChanPressNew.vi.
II. Run the VI by clicking the right arrow at the top of the screen. For
the calibration portion of the experiment, with the flow stopped,
click the 'Zero' button on the vi. It will take 2 additonal scanes to
zero the sensor offset. Save the zero pressure scan data to use in
correctiing for offset error later in your analysis.
III. Dial in the desired frequency using the Frequency Adjust knob on
the inverter while monitoring the Frequency indicator. Wait for 60
seconds for the flow to stabilize. Dial frequencies from 5 Hz – 45 Hz
going up by 5 Hz increments.
IV. Run the VI and record the values for Static Pressure (in m of water).
V. Use the formulas in the lab write-up to determine the velocities for
the frequencies from steps 4 and 5 and plot the velocities as a
function of frequency. Repeat steps 4 and 5 until pressure values for
all the frequencies have been recorded.
VI. Determine the frequencies that are required to operate the flow at
the desired velocities (0.6, 1.0 & 1.4 m/s) for the experiment by
interpolation of your data.

b. Cylinder Pressure Distribution.

I. Start the Vi
II. Enter your Zero offset values to correct for the system errors
III. Adjust the flow velocity to the speed desired by rotating the
Frequency Adjust knob on the inverter to the frequencies found in
step V & VI of the calibration section.
IV. Allow the flow to stabilize for 60 seconds and run the VI. The values
on the display may be stored to disk by entering a file name when
prompted.
V. Repeat steps III and IV until data for all three flow velocities.
6. For velocities plot the pressure coefficient as a function of angle around the
cylinder on a single graph. The pressure measurements are located at the
following angular positions.
No. Deg
1 0
2 22.5
3 45
4 56
5 67
6 78
7 90
8 101
9 112
10 123
11 135
12 157.5
13 180
7. Compare the measured pressure coefficient distributions to that for the
idealize flow around the cylinder Cp= 1-4 sin2θ, by plotting it along with the
measured pressure coefficient.
8. Using the analysis given in the discussion section, calculate the drag
coefficient for each of the three measured pressure coefficient distributions.
To check your integration procedure, also integrate the idealized in viscid
pressure coefficient distribution for a cylinder Cp= 1-4 sin2θ. The resulting
integral should give CDP = 0
9. Based on the error in the pressure measurements and the other uncertainties
in the parameters in the experiment calculate the error in the calculated drag
coefficients.
 Part 2: Water Tunnel Velocity Calibration and Drag Measurement
on a Cylinder From Laser Doppler Velocity Measurements

Objectives:

a) Re-Calibrate the velocity of the water tunnel as a function of impeller motor rpm
for use in subsequent hydrodynamic experiments using Laser Doppler velocity
MAE 171A117SA1126 Wn* T m d E x
m omWinter 2006
measurements RJ. C W S i mpdifisdby 0.U. nnan
b) Make measurements of the drag coefficient of a circular cylinder at three (0.6, 1.0
& 1.4 m/s) different flow velocities. Both velocity calibration and drag
measurements areCalibration
Water Tunnel Velocity based on andvelocity measurements
Drag Measurement on a from a laser-Doppler
Cylinder From Laser Doppler Velocity Measuremenh
Velocimeter.
Objective: (a) reconfirm the calibration of the velocity of the water tunnel at one or two flow
speeds to make sure the tlmnel operating characteristics have not changed. (b) Make
Discussion:
measurements of the drag coefficient of a circula~cyiinder at three different flow velocities.
Both velocity calibmtim and drag measurements are based on velocity measurements from a
laser-Doppler Velocimeter.
Laser Particle Velocimetry:
An
Laserunderstanding
Parlick Velocimetry:of the simplified theory of the operation of a laser-Doppler velocimeter
is necessary for performing the water tunnel velocity measurements.
An understanding of the simplified theory of the operation af a laser-Doppler velocimeter is
necessary for pdormiag the water tunnel velocity measurements.
When two monochromatic Laser beams cross, a fringe (interference) pattern is formed.
When two monchromatic Iaser bema cross, a fringe (interference) pattern is formed

MAE 171A1175A1126 WatmTmnd Bxpaimcot W i n t e r m

fig. I RJ. W l i e a , modihd by O J t Tynm

Due to the GuassimDueintensity dktdbutionintensity

to the Gaussian beam, theofftrnges
of the 1- distribution at the
the Laser beam, the focalat the center of
of fringes
centerthe
v o 1 w will be brighCer than at the edges. As a partide crosses the fringes, the w a
the focal volume will be brighter than at the edges. As a particle crosses h d lightthe fringes, the
wilI bwe the fouowing intensity si@:
scattered light will have the following intensity signal:

Intensity

Direction along Flow

Start End
start of
of Fringe Pattern end Direction a10
fringe pattern Flow

The faster the particle movee tbrough the focal volume, the higher the frequency of modulation
of intensity.
 start of end Direction a10
The faster the particle moves through
fringe pattern the focal volume, the higher the frequency of
Flow
modulation of intensity.
The faster the particle movee tbrough the focal volume, the higher the frequency of modulation
Theofvelocity
intensity. of the particle can be determined from the frequency times the fringe
spacing.
The velocity of the the fringe spacing.

The fringe spacing A is the product of the wavelength of the laser h divided by two times the sine
of the half angle between the two beams. @/2.
The Fringe Spacinghcan thus be calculated in terms of wavelength of Laser λ and angle
A=-
between the two beams
0 Θ as
2 sin
λ
Δ =For the argon-ion laser system wed in this e x m e n t
⎛ Θ ⎞
2sin⎜ ⎟
⎝ 2 ⎠

ForFor
thea 50
Argon-ion Laser
cm foal length lens system used in this experiment,

€ λ1 = 488.0nm = 4.88 × 10 −7 cm
λ2 = 514.5nm = 5.145 × 10 −7 cm

For a 50 cm focal length lens,

€ Θ = 2.32°
If the velocity in the direction of the flow is Ux, the frequency from a particle crossing the
fringe volume is
€
U x 2U x ⎛ Θ ⎞
f = = sin⎜ ⎟
Δ λ ⎝ 2 ⎠

The Laser-Doppler velocimeter instrument detects the optical signal scattered from
particles in the flow and determines the velocity from the measured frequency. By
€ traversing the Laser/detector head and the crossed laser beams across the water
channel, velocity profiles can be measured.
the flow and determines the velwity from the m e a d frequency. By traversing the
laser/detector head and the crossed laser beams otcrass the water c h e l velocity protiles can be
mead,

Finding the drw from the meamred velocity:

The calculation of the drag ccefficient of a cylinder examined the momentum loss in the wake
from the velocity
Finding measured with
the drag the LDV
from apparatus. The
the measured d y s i s is based on a control volume
velocity:
d y s i s using the geometry &own in the figure blow.
The calculation of the drag coefficient of a cylinder examined the momentum loss in the
wake from the velocity measured with the LDV apparatus. The analysis is based on a
control volume analysis using the geometry shown in the figure below.

The integral form of the mass conservation equation is given as:

The integral farm of the mass conservation equation is given as:
∫ ρU.d A = 0
Since there is no fluid flow through the top and bottom surfaces of the control volume
shown
Since there above,
is no fluid the
flow integral
through thebecomes
top and bottom surfaces of the control volume above, the
€integral becomes
2 3

∫ ρU dy − ∫ ρU dy = 0
1 2
1 4

Using the Using theabove,

geumetry geometry
this above,
integralthis integral
can be can be
rewriteen as rewritten as
h h
€
∫ U1dy − ∫ U 2dy = 0
0 0

And, since the upstream flow velocity is uniform this simplifies to give

€ h
U1∞ h − ∫ U dy = 0
2
0

Add and subtract U2 h where U2 is the free stream flow located downstream of the
∞ ∞

cylinder.
€
h
U1∞ h − U 2∞ h + ∫ (U 2∞ − U 2 ) dy = 0
o

A little algebra then gives

€
 h ⎛ U ⎞
(U1∞ − U 2∞ )h + U 2∞ ∫ ⎜1 − U 2 ⎟dy = 0
0 ⎝ 2∞ ⎠

If we define the displacement thickness, Δ as

h ⎛ U 2 ⎞
€ Δ= ∫ ⎝ U ⎟⎠dy
⎜1 −
0 2∞

We can then write a relationship between the free stream velocities upstream and down
stream of the object, the displacement thickness, and the height of the apparatus as
€
⎛ Δ ⎞
U1∞ = U 2∞ ⎜1 − ⎟
⎝ h ⎠ (1)

Eq.(1) can be used to calculate the change in the free stream velocity from state1
ahead of the cylinder to state 2 in the wake of the cylinder. However, unless boundary
€ layer effects at the water tunnel walls are taken into effect, this relationship will not be
very accurate.

The Drag force on the cylinder can be calculated from the momentum loss of the fluid
entering and exiting the control volume.

In general for a flow with uniform loss in flow velocity the drag would be given as
.
Drag = m .ΔU (2)

Where m is the mass flow rate around an object and U is the change in free stream flow
speed around this object. Again, this form assumes that the flow field is uniform in the
€ direction transverse to the mean flow vector.

Our situation is more complex, in that the flow magnitude does vary in the transverse
direction. In this case, the drag is given by momentum-balance arguments in terms of an
integral

Drag = ∫∫ ρU da.(U
2 1 − U2 )

The drag per length b of the cylinder is given by

€ h
Drag = b ∫ ρU 2 .(U1 − U 2 ) dy
0 (3)
Assuming the velocity ahead of the cylinder is uniform U1 = U1 and setting b=l, the Drag
∞

per unit length of cylinder is

€
h ⎛
U 2 ⎛ U 2 ⎞ ⎞
2
1
Drag = ρU1∞ ∫ 2⎜
2 ⎜ − ⎜ ⎟ ⎟⎟ dy
2 U
0 ⎝ 1∞ ⎝ U1∞ ⎠ ⎠
(4)

€
 With the assumption that U2 , = U1 ,
∞ ∞

1 h
U 2 ( y ) ⎛ ⎛ U 2 ( y ) ⎞⎞
2
Drag = ρU1∞
2
∫ U 2⎜⎜1 − ⎜ U ⎟⎟⎟dy
0 2∞ ⎝ ⎝ 2∞ ⎠⎠ (5)

The Drag coefficient CD is defined by normalizing the drag by the dynamic pressure
2
½ρU1 and the projected area of the cylinder Ap =D.(b =1) where D is the diameter of the
∞

€ cylinder and b=1is the unit length of the cylinder. The Drag coefficient per unit length of
cylinder is:

Drag Drag
CD = =
1 2 1 2
ρU1∞ .A p ρU1∞ .D
2 2 (6)

€
Experimental Setup:

The Laser Doppler system is used to make non-intrusive velocity measurements in a

fluid flow.

The system consists of an Solid State laser, a beam separator/shifter and splitter, a fiber
optics and lens system, a photomultiplier receiver, digital signal processors and a
computer system with the Flowsizer program installed. Flowsizer is the program that
performs numerical analysis on the received data from the signal processors and then
displays the results on the screen in addition to saving velocity information to
disk. The program also allows the selection of operating parameters for the hardware
and software for the system.
Experimental Procedure:

1. Turn the power to the computer on and click on the Flowsizer icon.
2. Turn the power on to the Unidex indexer that controls the traversing table.
3. Adjust the flow velocity to the speed desired by rotating the Frequency Adjust knob
on the inverter to the frequencies found in Part 1 for the desired Velocities.
4. Ensure the curtain around the experiment is closed such that a stray beam or reflection from
the laser cannot escape and pose a hazard to other personnel in the laboratory. Ensure also
that all members of the group are wearing their safety glasses.

5. Turn on the FSA 3500 (Power switch is located on the rear panel)
6. Turn on the Powersight laser system. (Power switch is located on the rear panel)
Wait 1 minute to let the power supply energize the laser
Turn the laser key to the on position.
Press the silver button to output the laser beam, a red light should illuminate around the button.
7. Make sure the aperature on the laser is in the fully open position and particles are
injected into the channel.
8. Follow the instructions below to run the calibration and the velocity measurement
experiment.
a. Calibration
I. At the FlowSizer console, under Home category -> Run management panel,
select MAE in the Current drop down box.
II. Hardware category -> Traverse panel -> Start Travese Manager
III. In Manual tab, press on Abs Home to move the traverse to its hardware zero reference
point. Wait until the hardware finish moving, verify visually that it is at the home position
(the laser should be at a location downstream of the cylinder).This puts the traverse at a
known reference location, indicated as 0,0,0. All future indicated positions are
displacements from this point.
IV. The top lines in X, Y, Z, group boxes show the current positions of the
x,y, and z axies of the traverse (z axis is not used in this experiment).
The boxes next to the Move button allow you to enter a new position for
each axis.
Enter 23.5 and 5.25 to the X and Y boxes next to the Move button
correspondingly and press Move.
V. Dial in the desired frequency using the Frequency Adjust knob on the inverter while
monitoring the Frequency indicator. Wait for 60 seconds for the flow to stabilize. Dial
frequencies from 5 Hz - 45 Hz going up by 5 Hz increments.
VI. Home category -> Run Capture panel -> press Begin.
This will start the data capturing process.
VI. Home category -> Run Management panel -> press Save.
Enter a name for the run and save it.
VII. To play back or export the data:
Home category -> Run management panel -> select the name you entered
in step VI in the Current drop down box -> Hit Playback in Run Capture
panel -> Press Data Sets to CSV file to export data (Notice: the X & Y
position can't be read).
VIII. Continue to increment the Frequency on the motor inverter and repeat step VI
and VII until the velocity for the maximum frequency has been recorded
 IX. Plot the frequencies and velocities. Determine the frequencies that correspond
to the three velocities (0.6, 1.0 & 1.4 m/s) that will be used to take actual data
for determining drag of the cylinder.
X. Use the calibration values of Frequency that correspond to the desired
velocities for the remainder of the experiment.
b. Velocity Measurements
I. There are four sets of data that will be taken at three different velocities for this
experiment. There is one axial set that provides velocity information for various
centerline positions in the tunnel 6 inches upstream and 8 and 16 inches
downstream of the cylinder. In addition, there are three sets of data that
provide velocity information across the water tunnel (transverse) at 6 inches
upstream and 8 and 16 inches downstream of the cylinder.
II. To begin taking data, dial the Frequency that corresponds to the velocity
desired into te inverter. Allow 60 seconds for the flow to stabilize.
III. At the FlowSizer console, under Home category -> Run management panel,
select MAE in the Current drop down box.
IV. Hardware category -> Traverse panel -> Start Travese Manager
V. In Manual tab, press on Abs Home to move the traverse to its hardware zero reference point.
Wait until the hardware finish moving, verify visually that it is at the home position (the laser
should be at a location downstream of the cylinder).This puts the traverse at a known
reference location, indicated as 0,0,0. All future indicated positions are displacements from
this point.
VI. In Matrix tab, press Open and select Traverse 1.TRV. It is the transverse
position matrix for location X = 6 inches upstream of the cylinder. Make
sure the "MAE" is used under Run Setup column.
VII. Hardware category -> Traverse panel -> Start Scan Capture.
A window will pop up and shows the current location of data
capture. The data is saved to Location0000xx. Wait few
mintues for the whole process to finish.
VIII. To play back or export the data:
Home category -> Run management panel -> select Location0000xx in the
Current drop down box -> Hit Playback in Run Capture panel -> Press Data
Sets to CSV file to export data (Notice: the X & Y position can't be export,
but you have it in your position matrix in step VI). Repeat this step until all
your data captured in VII is exported.
VII. Repeat step III to VIII and using Traverse 2.TRV, Traverse 3.TRV
and Axial.TRV in step VI. Repeat this step with a different desired flow
velocity
XII. Turn the Frequency knob fully CCW at the end of the experiment. The
laboratory assistant or TA will power the system down at the end of the lab
session
10. Plot the velocity and the RMS velocity for each flow condition as a function of
distance from the cylinder on separate graphs and compare with the velocity from
the pitot probe calibration. Also, on a separate graph plot the RMS velocity
normalized by the absolute value of the local mean velocity as a function of distance
from the cylinder. Is there an upstream influence of the flow around the cylinder?
11. Calculate the drag coefficient for the cylinder for 3 velocities using the velocity
profiles obtained in the earlier steps and equation 5 given in the Discussion session.
Compare the measured ratio of U2 / U1 with that calculated from Eq. 1. You are
∞ ∞

required to calculate the error in the Drag coefficient based on the error in the
velocity measurements.
12. Compare the drag coefficient CD determined from the wake velocity measurements
with the pressure drag coefficient CDP, in a table as a function of the Reynolds
number of the flow based on the diameter of the cylinder.
13. Compare these drag coefficient results with those obtained using the pressure
distribution method used in part 1, and also compare both sets of results with those
found in any fluid mechanics textbook.

- Chi Wai Chow

January - 2014

References:
1. John D. Anderson, “Fundamentals of Aerodynamics” 2nd Ed., McGraw Hill, 1991
2. L. E. Drain, “The Laser Doppler Technique”, John Wiley & Sons, 1980
3. R. J. Cattolica, G. R. Tynan, “Water Tunnel Experiment”, 2006