WIND TUNNEL Notes

Wind Tunnel Techniques
3003
Wind tunnel techniques course depicts the types, working and characteristics of wind tunnels in
the laboratory. The flow characteristics and flow visualization in the tunnel are recorded for
further observations.
Objectives
The course should enable the students to:

1) Understand the non-dimensional number by Buckingham theorem
2) Differentiate the wind tunnels on the basis of circuit, air flow and working..
3) Know the calibration of a wind tunnel.
4) Understand the pressure and force measurements in wind tunnel
5) Deduce the flow visualization techniques used in the wind tunnel testing
Outcome
The students should be able to:
1) Solve the Buckingham theorem to find the SI unit of a parameter
2) Clearly understand the working of blow down, in draft tunnels and their specifications
3) Know about horizontal buoyancy, flow angularities are checked while calibration
4) Know about component axis balance and internal balances are read and understood for the
measurements in wind tunnel
5) Get a clear idea about the smoke and tuft flow visualization procedures in WT testing
AE 2751
July-Nov. 2016
Syllabus
Unit
I
Topic
No. of
Lectures
Principles of Model Testing
Buckingham theorem, Non-dimensional numbers, Scale effects, Types of similarities
II
Wind tunnels
Classification, Special problems of testing in subsonic, transonic, supersonic and

hypersonic speed regions, Layouts, Sizing and design considerations
III
Calibration of Wind Tunnels
11
Test section speed, Horizontal buoyancy, Flow angularities, Turbulence

measurements, Associated instrumentation, Calibration of supersonic tunnels
IV
Wind Tunnel Measurements
12
Pressure and velocity measurements, Force measurements, Three and six

component balances, Internal balances
AE 2751
July-Nov. 2016
Syllabus
Unit
V
Topic
No. of
Lectures
Flow visualization techniques
Surface and turft flow visualization techniques, Dye injection techniques, Optical
methods of flow visualization
References
1 Barlow, Jewel B. Rae, William H. and Pope, Alan Low speed wind tunnel testing, III Ed., ISBN 13: 9780471557746, ISBN 10:
0471557749, 1999 (Wiley India Edition, 2010, About Rs. 1,000).
2 Pope, Alan and Goin, Kenneth L, High-Speed Wind Tunnel Testing, ISBN-13: 978-0882757278; ISBN-10: 088275727X, 1978.
Instructor:
Test
Dr. N. Sitaram
E_mail: pt.nsitaram@hindustanuniv.ac.in
First Periodical
Second
Model
Seminar/
Attendance End-semester
Test*
Periodical Test* Exam Assignment/Quiz
Examination
Weightage
Duration
10%
2 periods
AE 2751
10%
2 periods
20%
3 hours
10%
-
10%
-
50%
3 hours
July-Nov. 2016
Introduction
CFD vs. EFD
Computational Fluid Dynamics
Requirements High end computer, software, printers,
plotters etc.
Experimental Fluid Dynamics

Experimental facility (Wind tunnel),
Instrumentation etc.
Manpower
One or two for small CFD work,

A small team for large CFD work, each
member specializing in grid developing,
solver, post processing etc.
Usually requires moderate number of

people for small EFD for preparing and
installation of model, for measurements etc.
Requires large number of people with
different skills for large experiments.
Initial Cost
Moderate
High
Time required
Moderate for computation, but large for

post processing
Large for preparation, small for data

acquisition and post processing
Information
acquired
Hugh
Limited
CFD and EFD are complimentary. Both require careful working. Needs verification and
validation, Extensive CFD can be carried out followed by limited comparison with EFD
AE 2751
July-Nov. 2016

Buckingham theorem
Dimension is characteristic of the object, condition, or event and is described
quantitatively in terms of defined units.
A physical quantity is equal to the product of two elements:
A quality or dimension
A quantity expressed in terms of units
Dimensions
Physical things are measurable in terms of three primitive qualities (Maxwell 1871)
Mass (M)
Length (L)
Time (T)
NOTE: Temperature, electrical charge, chemical quantity, and luminosity were added as
primitives some years later.)
AE 2751
July-Nov. 2016

Buckingham theorem
Examples
Length (L)
Velocity (L/T)
Force (ML/T2)
Units:
Measurements systems: CGS, MKS, SI
SI units are now the international standard (although many engineers
continue to use Imperial or U.S.)
AE 2751
July-Nov. 2016

Buckingham theorem
SI Primitive Units
Dimension
Symbol
Unit
Symbol
Length
meter
Mass
kilogram kg
Time
second
Temperature
Kelvin
Elec. Current
Ampere
Luminous intensity
Candela cd
Amount of substance
Mole
AE 2751
mol
July-Nov. 2016

Buckingham theorem
SI Derived units
Description Derived unit Symbol Dimension
AE 2751
Force
Newton
kg m/s2
Energy
Joule
kg m2/s2
Pressure
Pascal
Pa
kg/(ms2)
Power
Watt
kg m2/s3
July-Nov. 2016

Buckingham theorem
Fundamental rules:
All terms in an equation must reduce to identical primitive dimensions
Dimensions can be algebraically manipulated.
Uses:
Check consistency of equations
Deduce expression for physical phenomenon
AE 2751
July-Nov. 2016

Buckingham theorem
Simple Example: Drag on a Sphere
Drag depends on FOUR parameters:

sphere size (D); velocity (V); fluid density (); fluid viscosity ()
Difficult to know how to set up experiments to determine dependencies
Difficult to know how to present results (four graphs?)
AE 2751
July-Nov. 2016

Buckingham theorem
F
=
V D
Only one dependent and one independent variable
Easy to set up experiments to determine dependency
Easy to present results (one graph)
AE 2751
July-Nov. 2016

Buckingham theorem
Experimentally derived relation between the dependent and independent variables
AE 2751
July-Nov. 2016

Buckingham theorem
Step 1:
List all the parameters involved
Let n be the number of parameters
Example: For drag on a sphere, F, V, D,
Hence
n=5
Step 2:
Select a set of primary dimensions
For example M (kg), L (m), T (s).
Example: For drag on a sphere, choose MLT
AE 2751
July-Nov. 2016

Buckingham theorem
Step 3
List the dimensions of all parameters
Let r be the number of primary dimensions
Example: For drag on a sphere r = 3
Parameter:
Unit:
ML/T
V
2
L/T
D
2
M/L
M/LT
Step 4
Select a set of m dimensional parameters that includes all the primary dimensions
Example: For drag on a sphere (m = r = 3) select , V, D
AE 2751
July-Nov. 2016

Buckingham theorem
Step 5
Set up dimensionless groups ps
There will be (n m)=2 equations
Example: For drag on a sphere
ML M
T L
M LT =
= F V D
L
T
Exponents to be determined to satisfy dimensional homogeneity

Equating exponents of M, L and T on left and right sides
For M: 0=1+a
Hence a=-1
For T: 0=-2-b
Hence
For L: 0=1-3a+b+c
0=1+3-2+c
b=-2
Hence
c=-2
Hence the first dimensionless number can be written as
AE 2751
F
V D
July-Nov. 2016

Buckingham theorem
Step 6
Check to see that each group obtained is dimensionless
Example: For drag on a sphere
AE 2751
F
V D
ML L
T M
M
L
T
L
1
L
L
LT
L
T
M
July-Nov. 2016

Buckingham theorem
Direct Rationalization of Dimensionless Groups
Obtain the terms by simple reasoning:
1
Determine their number from Buckingham theorem. Make sure the appropriate
number of groups is obtained.
Identify variables (g etc.) that contain force quantities and formulate the
2 2
corresponding force. For example, F=V L and F=VL. Then take ratios of
these forces (F/F=VL/) to get dimensionless groups.
3
If the force groups do not comprise the number of dimensionless groups sought,
then look for length type terms, velocity type terms, and/ or time type variables
which can be divided to give dimensionless groups.
AE 2751
July-Nov. 2016

Buckingham theorem
Direct Rationalization of Dimensionless Groups
There may be more than one length in the problem, i. e. L and D which give the
dimensionless group, L/D. Length type variables can also include area and volume, so
That A/L2 and V/L3 are dimensionless groups.
Velocity type terms include velocity, V, angular velocity,
w, and volume flow rate, Q,
which yield dimensionless groups, wr/V and Q/VL .

2
Time type terms consist of period, , of the motion and convective time scale, L/V, which give
the dimensionless group, V/L
AE 2751
July-Nov. 2016

Types of forces encountered in fluid phenomenon
Inertia Force, Fi: = mass X acceleration in the flowing fluid.
Viscous Force, Fv: = shear stress due to viscosity X surface area of flow.
Gravity Force, Fg: = mass X acceleration due to gravity.
Pressure Force, Fp: = pressure intensity X C.S. area of flowing fluid.
AE 2751
July-Nov. 2016

Significant Dimensionless Groups in Fluid Mechanics
These are numbers which are obtained by dividing the inertia force by
viscous force or gravity force or pressure force or surface tension force
or elastic force.
As this is ratio of once force to other, it will be a dimensionless number.
These are also called nondimensional parameters.
The following are most important dimensionless numbers.
Reynolds Number
Froude Number
Euler Number
Mach Number
These numbers are named after Fluid Mechanics Scientists who had
established their significance.
AE 2751
July-Nov. 2016
10

Reynolds Number, Re: It is the ratio of inertia force to the viscous force
of flowing fluid.
Velocity
Volume
Mass.
. Velocity
Fi
Time
Time
Re
Fv Shear Stress. Area Shear Stress. Area

Q.V AV .V AV .V VL VL
du
V
.A
.A .A
dy
L
AE 2751
July-Nov. 2016

Euler Number, Eu: It is the ratio of inertia force to the pressure force of
flowing fluid.
Velocity
Fi
Time
Fp
Pr essure. Area
Mass.
Eu
AE 2751
Q.V
P. A
AV .V
P. A
V2
P/
Volume
. Velocity
Time
Pr essure. Area
V
P/
July-Nov. 2016
11

Froude Number, Fe: It is the ratio of inertia force to the gravity force of
flowing fluid.
Velocity
Fi
Time
Fg
Mass. Gavitational Acceleraion
Mass.
Fe
Volume
. Velocity
Time
Mass. Gavitational Acceleraion
AV .V
V2
V
Volume.g
AL.g
gL
gL
Q.V
AE 2751
July-Nov. 2016

Mach Number, M: It is the ratio of inertia force to the elastic force of
flowing fluid.
Fi
Fe
Q.V
K .A
Where : C
AE 2751
Velocity
Time
Elastic Stress. Area

Mass.
AV .V
K .A
L2V 2
2
KL
Volume
. Velocity
Time
Elastic Stress. Area
V
V
K / C
K /
July-Nov. 2016
12

Dimensionless numbers for force quantities
Force intensities, pressure and shear stresses, are non-dimensionalized with the dynamic pressure
Pressure coefficient:
Local shear stress coefficient:
Resultant forces, such as lift and drag, are non-dimensionalized with the dynamic pressure force,
V A, where A is some characteristic area.

Lift coefficient, C =
AE 2751
and
Drag coefficient, C =
July-Nov. 2016

Dimensionless numbers for force quantities
Torque, T, is force F times a moment arm, L, which is non dimensionalized by multiplying the
dynamic pressure force by L to give torque coefficient.
Torque coefficient:
C =
Power is the rate of doing work, FL/T or FV, which requires the dynamic pressure force be
multiplied by V to give power coefficient.
Power coefficient:
AE 2751
C =
July-Nov. 2016
13

Elaborate Example: Power required to rotate a propeller
The power, P required to rotate a propeller depends on the fluid density, , the propeller
diameter, D, the rotational speed, w, the velocity, V, of the fluid approaching the
propeller, the speed of the sound, C, and the angle, a, of a propeller blade.
P = f (, D, w, V, C, a)
Seven variables (n=7) give four dimensionless groups (m=n-r=4).
Two force like terms yield one dimensionless group.
The dynamic pressure, can be used to non-dimensionalized forces. As power is F V,
P
1
V A
2
AE 2751
July-Nov. 2016

For a rotating propeller, L=D and V=wD to obtain
=C =
P
1
D
2
There are three velocity variables:
w, V and C, which give two dimensionless numbers.
V
V
= Advace ratio =
= Mach number =
D
C
AE 2751
July-Nov. 2016
14

Need one more dimensionless group, but have not yet considered the angle, a. An angle can
be defined in terms of its tangent which is the ratio of two lengths and is dimensionless.
That is a is a
term.
=
Then
C =
=
or
AE 2751
P
1
2 D
V V
, ,
D C
, , ,
July-Nov. 2016

Dimensional Analysis
Definition: Dimensional analysis is a process of formulating fluid mechanics problems in terms
of non-dimensional variables and parameters.
Why is it used :
Reduction in variables [ If F (A1, A2, , An) = 0, then f (P1, P2, Pr < n) = 0,

where, F = functional form, Ai = dimensional variables, Pj = non-dimensional
parameters, m = number of important dimensions, n = number of dimensional variables,
r=nm). Thereby the number of experiments required to determine f vs. F is reduced.
Helps in understanding physics

Useful in data analysis and modeling
Enables scaling of different physical dimensions and fluid properties
AE 2751
July-Nov. 2016
15

Dimensional Analysis: Example
Drag = f (V, L, , C, t, e, T, etc.)

From dimensional analysis,
Vortex shedding behind cylinder
AE 2751
July-Nov. 2016

Similarity and Model Testing
Definition : Flow conditions for a model test are completely similar if all relevant
dimensionless parameters have the same corresponding values for model and prototype.
i model = i prototype for i = 1 to n

Enables extrapolation from model to full scale
However, complete similarity usually not possible. Therefore, often it is necessary to
use Re, or Fr, or Ma scaling, i.e., select most important parameter and
accommodate others as best possible.
AE 2751
July-Nov. 2016
16

Dimensional Analysis and Similarity
Geometric Similarity: The model must be the same shape as the prototype. Each
dimension must be scaled by the same factor.
Kinematic Similarity: Velocity as any point in the model must be proportional

Dynamic Similarity: All forces in the model flow scale by a constant factor to
corresponding forces in the prototype flow.
Complete Similarity: is achieved only if all above three conditions are met.
AE 2751
July-Nov. 2016

Dimensional Analysis and Similarity
Complete similarity is ensured if all independent groups are the
same between model and prototype.
What is ?
We let uppercase Greek letter denote a nondimensional parameter,
e.g., Reynolds number Re, Froude number Fr, Drag coefficient, CD, etc.
Consider automobile experiment
Drag force is F = f (V, , L)
Through dimensional analysis, we can reduce the problem to
AE 2751
July-Nov. 2016
17

Flow Similarity and Model Studies
Example: Drag on a Sphere
AE 2751
July-Nov. 2016

Example: Drag on a Sphere
For dynamic similarity
.then .
AE 2751
July-Nov. 2016
18

Scaling with Multiple Dependent Parameters
Example: Centrifugal Pump
Pump head:
h=g1(Q, wD, )
Pump power:
P=g2(Q, wD, )
AE 2751
July-Nov. 2016

Similitude-Type of Similarities
Geometric Similarity: is the similarity of shape.
Lp
Lm
Bp
Bm
Dp
Dm
Lr
Where: Lp, Bp and Dp are Length, Breadth, and Diameter of Prototype

and Lm, Bm, Dm are Length, Breadth, and Diameter of Model.
LR= Scale ratio
AE 2751
July-Nov. 2016
19

Kinematic Similarity: is the similarity of motion.
V p1
Vm1
Vp 2
Vm 2
Vr ;
a p1
am1
ap2
am 2
ar
Where: vp1& vp2 and ap1 & ap2 are velocity and accelerations at point 1
& 2 in prototype and vm1 & vm2 and am1 & am2 are velocity and
accelerations at point 1 & 2 in model.
Vr and ar are the velocity ratio and acceleration ratio
AE 2751
July-Nov. 2016

Dynamic Similarity: is the similarity of forces.
Fi p Fv p Fg p
Fi m Fv m Fg m
Fr
Where: (Fi)p, (Fv)p and (Fg)p are Inertia, Viscous and

Gravitational Forces in Prototype and (Fi)m, (Fv)m And (Fg)m are
Inertia, Viscous and Gravitational Forces in Model.
Fr Is The Force Ratio
AE 2751
July-Nov. 2016
20

Example: Centrifugal Pump
Head coefficient:
Power coefficient:
AE 2751
July-Nov. 2016

Example: Centrifugal Pump (Negligible viscous effects)
If
then
AE 2751
then neglecting Reynolds number (viscous effects)
gh
gh
=
D
D
and
July-Nov. 2016
21
Why Wind Tunnel Testing?

Consider an aircraft flying in still or almost still atmosphere.
It is possible to get the aircraft performance from flight tests, but complicated and resource
(time, money, human resources etc.) consuming. Also only limited information can be obtained.
Alternatively a still aircraft model can be tested in a wind tunnel with air flowing at a speed of U.
This reduces the resources required and more information can be obtained.
A wind tunnel can be used to test models of aircraft and other vehicles and components of
aircraft (ex. wings).
Alternative testing resources: Whirling arms (no longer used), Water tunnels
Can CFD replace EFD (wind tunnels)?
AE 2751
July-Nov. 2016
CFD vs. EFD

Computational Fluid Dynamics
Experimental Fluid Dynamics
Requirements
High end computer, software, printers,

plotters etc.
Experimental facility (Wind tunnel),

Instrumentation etc.
Manpower
One or two for small CFD work,

A small team for large CFD work, each
member specializing in grid developing,
solver, post processing etc.
Usually requires moderate number of

people for small EFD for preparing and
installation of model, for measurements etc.
Requires large number of people with
different skills for large experiments.
Initial Cost
Time required
Moderate
High
Moderate for computation, but large for

post processing
Large for preparation, small for data

acquisition and post processing
Information
acquired
Hugh
Limited
CFD and EFD are complimentary. Both require careful working. Needs verification and validation,
Extensive CFD can be carried out followed by limited comparison with EFD
AE 2751
July-Nov. 2016
National Aerospace Laboratory Trisonic Wind Tunnel Facility
Test section: 1.2mx1.2m
Operation: Intermittent blowdown
Mach Number Range: 0.2 to 4.0
AE 2751
Test duration: 40 secs (Typical)
Reynolds number: 8 to 60x106 per meter
July-Nov. 2016
Lockheed Martin High Speed Wind Tunnel Circuit Layout

Mach Number Range:
Transonic range:
0.3 to 1.8
Supersonic range: 1.6 to 4.8
Reynolds Number Range

4 to 34x106/foot
Equivalent Airspeed
150 to 1,200 knot
(280 to 2200 kmph)
(144 to 1132 mps)
Run Time: 15 to 110 secs

AE 2751
July-Nov. 2016
NASA Ames Wind Tunnel

World largest wind tunnel
Test sections: 40x80& 80x120
Speeds:
300 knots 100 knots

556 kmph 185 kmph
154 mps 51 mps
80x120 test section can test full size

Boeing 737 aircraft at velocities up to
unit Reynolds numbers of 1.1x106/ft.
AE 2751
July-Nov. 2016
F/A-18 Fighter Aircraft Testing in NASA Ames Wind Tunnel
AE 2751
July-Nov. 2016
Classification of Wind Tunnels

I
Based on Size:
Small: Test section 0.3 mx0.3 m, mainly for college instructional purposes
Medium: Test section 1 mx1 m, used for university research purposes
Large: Test section 2 mx2 m, used for testing industries and research laboratories
Very Large: Test section 4 mx4 m, used for testing large models in industries and research
laboratories (for ex: NASA Ames Research Laboratory)
II Based on Wind Velocity:

Low velocity:
Velocity less than 100 m/s, incompressible flow used in instructional and
university research purposes
Subsonic velocity: Compressible flow (M<0.8) used for university, industry and research
laboratory research purposes. No shock waves occur.
Transonic velocity: Compressible flow (M 0.8 to 1.2) used for university, industry and
research laboratory research purposes. Shock waves occur.
Supersonic velocity: Compressible flow (M=1.2 to 5) used for university, industry and
research laboratory research purposes. Shock waves occur.
Hypersonic velocity: Very high Mach numbers (M>5) used for industry and research
laboratory research purposes. Rarefied gas with very low density.
AE 2751
July-Nov. 2016
Classification of Wind Tunnels

III Based on Test Section Configuration:
Open Circuit: Used mainly for small and intermediate tunnels. Test section may be
without solid boundaries (open jet) or with solid boundaries (closed jet)
Closed Circuit: Used mainly for large and very large tunnels, may have closed or open
(Fluid Mechanics Laboratory, Department of Applied Mechanics, IIT Madras). May have
single return, double return or annular return.
IV Based on Operation:
Continuous:
Most of the tunnels operate continuously.
Short Duration: These tunnels usually operate for very small time (Intermittent: A few
minutes, Blow down: a few seconds) These types of tunnels are used in high speed
turbomachinery testing. Saves enormous amount of energy for testing.
V Special Tunnels:
High Reynolds Number Tunnels
Low Turbulence Tunnels: The tunnel test section usually have a turbulence level of 1%.
Certain applications require very low turbulence levels 0.1% (MTL wind-tunnel at the
Department of Mechanics, KTH, Sweden)
Variable Density Tunnels: Independent variation of Mach and Reynolds numbers.
Cryogenic tunnels.
AE 2751
July-Nov. 2016
Applications of Wind Tunnels

I
Aeronautical applications
Most of wind tunnels are used for aeronautical applications for instructional purpose, for
research carried out at universities, industries and research laboratories. The research
carried out may be for fundamental understanding of fluid flow phenomena or for
developmental testing of components for aircrafts. Even moderate sized models of actual
aircraft are tested in large wind tunnels.
Aeronautical wind tunnels can be further classified as high Reynolds number wind tunnels,
V/STOL wind tunnels, Free-flight wind tunnels, Spin tunnels or vertical wind tunnels,
Stability tunnels, Propeller tunnels, Propulsion tunnels, Icing tunnels, Low turbulence
tunnels, Two-dimensional tunnels.
II Smoke tunnels
III Automobile wind tunnels
IV Aeroacoustic wind tunnels (Anechoic tunnels)
AE 2751
July-Nov. 2016

V Water tunnels
Used for flow visualization studies, underwater vehicle development and for cavitation
studies. Usually small size and works at low velocities to obtain realistic Reynolds
numbers.
III General purpose wind tunnels

Used for study of people drag (bike racers, skiers etc.), birds and insects, wind power
devices solar collectors, radar antennas and satellite television receivers, sails and
above-water parts of ships, bridges, etc.
IV Environmental or meteorological wind tunnels

These tunnels are designed to simulate Earths natural boundary layer, which typically
has a thickness of 300 to 600 m. They are used for study of static loads and associated
experiments on buildings, dynamic loads on buildings, unsteady aerodynamics in wind
engineering, agricultural/wind breaks, agricultural/agronomy, agricultural/soil erosion
breaks, snow drifting, evaporation and related issues, pollution dispersion, etc.
The reader is referred to Barlow J. B., Rae Jr. W. H. and Pope A., Low Speed Wind Tunnel
Testing, IIIrd Edition, John Wiley & Sons, Inc.,1999 for some more details of the above wind
tunnels. Indian edition of this text book is also available at about Rs. 1,000/-.
AE 2751
July-Nov. 2016
Snowdrift study for lodge, near building in photo, at a ski resort.

Note the effect of tower at left rear, which provides self-removal
of snow from entrance with prevailing from left.
AE 2751
Skier in wind tunnel
July-Nov. 2016
Effects of wind barriers on solar collectors

being simulated in a wind tunnel
Above: Efflux velocity is equal to wind speed and
slack height is 1.5 times building height.
Building is downstream of stack.
Below: Same conditions as above except building is
upstream of the stack simulated in a wind tunnel
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Open Circuit Wind Tunnel
Plan view of an open circuit wind tunnel (Daimler-Benz Aerospace Airbus, Bremen, Germany)
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

The following are the advantages and disadvantages of an open circuit wind tunnel
Advantages
1) Construction cost is typically much less.
2) If one intends to run internal combustion engines or do extensive flow visualization via.
smoke, there is no purging problem provided both inlet and exhaust are open to the
atmosphere.
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Disadvantages
1) If located in a room, depending on the size of the tunnel to the room size, it may require
extensive screening at the inlet to get high quality flow. The same may be true, if the inlet and/
or exhaust is open to the atmosphere, when wind and cold/hot weather can effect operation.
2) For a given size and speed, the tunnel will require more power to operate. This is usually a
factor only if used for development experiments, where the tunnel has high utilization rate.
3) In general, open circuit wind tunnels tend to be noisy. For larger tunnels (test sections of
6 m2 and more), noise may cause environ mental problems, limit hours of operation, and/or
require extensive noise treatment of the tunnel and surrounding room.
Because of low initial cost, an open circuit wind tunnel is often ideal colleges and universities,
where a tunnel is required for class room instruction purposes and research and high utilization
Is not required. Open circuit design are also frequently used by science fair participants who
build their own wind tunnels.
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Closed Circuit Wind Tunnel
Plan view of a closed circuit wind tunnel (Defense Establishment Research Agency,
DERA, 13x9 ft (3.9x2.7 m=10.9 m2) tunnel in Bedford, England
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Instrumentation
stand
Open test
section: 1 m dia.
Closed circuit (return flow) open test section wind tunnel,

Fluid Mechanics Laboratory, Department of Applied Mechanics, IIT Madras
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

The following are the advantages and disadvantages of a closed circuit wind tunnel
Advantages
1) Through the use of corner turning vanes and screens, the quality of the flow can be well
controlled and most important will be independent of other activities and in the building
and weather conditions.
2) Less energy is required for a given test-section size and velocity. This can be important
for a tunnel used for developmental experiments with high utilization (two or three shifts,
five to six days a week).
3) There is less environmental noise when operating.
Disadvantages
1) The initial cost is higher due to return ducts and corner vanes.
2) If used extensively for smoke visualization experiments or running of internal combustion
engines, there must be a way to purge tunnel.
3) If tunnel has high utilization, it may have to have an air exchanger or some other method
of cooling.
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Open or Closed Test Section?
An open test section in conjunction with an open circuit wind tunnel requires an enclosure
around the test section to prevent air being drawn into the tunnel from the test section rather
than the tunnel inlet.
For closed return wind tunnels of large size with an external balance, the open test section
tends to have one solid boundary, since the balance must be shielded from the wind.
This is an anomaly for aircraft experiments other than takeoff and landing, but it is a natural
condition for experiments on automobiles or surface-borne marine vehicles.
Many open test-section, closed circuit wind tunnels have experienced severe flow fluctuation
problems that require extensive post construction diagnostics and corrective actions.
One of the tunnels currently acknowledged to be one of the most useful tunnels in existence
nevertheless has had substantial difficulties with unsteady flow and noise when running in
the open test-section configuration.
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Open or Closed Test Section?
The most common geometry is a closed test section, but a wide range of tunnel geometries
have provided good experimental conditions, once the tunnel idiosyncrasies have become
known to the operators and users.
Slotted wall test sections are becoming more common as are test sections that can be
converted among two or more configurations.
A rectangular test section is preferable in larger size wind tunnels, as it is easier to change a
model when working off a flat surface.
Further, if automobile or other ground vehicle experiments are to be conducted, a flat floor
is a requirement.
AE 2751
July-Nov. 2016
10
Wind Tunnel Design

General Layout: Open Circuit and Blower Tunnel
Open Circuit: This type of
tunnel is widely used for
instructional purposes and for
investigations of fundamental
flow phenomena. The test
section is closed type (a).
Blower Tunnel: This type of

tunnel also is widely used for
calibration of flow devices. A
characteristic design problem
for these facilities is choice of
details of a wide-angle diffuser
between the fan or blower and
the settling chamber before
the contraction. The test
section is open type (a).
AE 2751
July-Nov. 2016
Wind Tunnel Layouts

Test Section Size
In general it may be expected that the test section should have as large a cross-sectional
area as possible.
Ideally, a tunnel would be large enough to handle a full-scale vehicle (aircraft, car etc.).
In fact, several tunnels were built in 1920s and through the 1940s to achieve this goal for aircraft.
However, since World War II era, and presumably in the future, the size of the aircraft have
become such that wind tunnels to accommodate full-scale vehicles are not practical.
If one uses the rule of the thumb that the model span should be less than 0.8 of the tunnel
width, then Howard Hughes Hercules, or as more popularly known, the Spruce Goose,
which was designed and built in the 1940s with a 320 ft wing span, would require a test
section 400 ft wide.
The cost of building and operating a tunnel of this size is staggering to contemplate.
AE 2751
July-Nov. 2016
11
Wind Tunnel Layouts

Test Section Size
The cost of building a model, transporting it, and erecting in the tunnel, as well as making
changes during an experimental program, would also be an interesting, albeit expensive task.
Thus, it is apparent for the larger of modern aircraft, the concept of a tunnel to accommodate
full-size aircraft is out of question based on costs and practical difficulties.
Recalling from the earlier discussions of flow similarity, it is more important to seek to obtain
Reynolds numbers for the model experiments that are as nearly as possible to the full-scale
values than to be concerned with size alone.
In practice most development are done in tunnels with widths from 10 to 20 ft (3 to 6 m).
AE 2751
July-Nov. 2016
Wind Tunnel Design

Basic Considerations
The first step in the design of a wind tunnel is to determine the size (small, intermediate,
large or very large) and shape of the test section (square, octagonal, circular or
rectangular), based on the intended uses of the tunnel.
The details given address tunnels for which the primary use is vehicle and vehicle
components.
A major part of the testing will be force testing, where information is sought for
performance, fuel efficiency, stability or control of a vehicle that may be an aircraft, an
automobile, a racing car, a submarine, a racing yacht, or possibly others..
The cross-sectional area of the test section basically determines the overall size of the
wind tunnel.
The size of the wind tunnel will be the primary factor in determining the structural or
shell costs, and the power and operating hours will determine the energy portion of the
operational cost.
AE 2751
July-Nov. 2016
12
Wind Tunnel Design

Basic Considerations
The major cost of operation is the salaries of the tunnel personnel.
But the electrical energy cost to run the tunnel and its auxiliaries is not an insignificant
cost and it will increases in the long run.
The details given address tunnels for which the primary use is vehicle and vehicle
components.
There must be a balance between initial costs and operating cost.
In the past many tunnels have been built with short diffusers and related features to
allow short circuit length to hold down initial cost while accepting higher energy costs
of operation.
This trade-off should be carefully examined with due consideration given to

anticipated energy costs, which are likely to increase.
AE 2751
July-Nov. 2016
Wind Tunnel Design

Overall Aerodynamic Objective
The overall aerodynamic objective of most wind tunnels is to obtain a flow in the test
section that is as near as possible to a parallel steady flow with uniform velocity
throughout the test section.
Perfection is not possible so every design is bounded by constraints that include

maximum cost, available space, available time, and available knowledge.
It is almost always desired to obtain the largest size of the test section and the
highest speed for the available funds.
High speed and large size are of course competing demands.
AE 2751
July-Nov. 2016
13
Wind Tunnel Design

Basic Decisions: Unit Reynolds Number
A central issue in the sizing of a low-speed wind tunnel is the achievable Reynolds
numbers for the models that can be accommodated.
The same question arise for vehicle developers who must select from available wind
tunnels one which to carry out the tests in a development program.
Reynolds number, Re=r V L/m .
A maximum Mach number of 0.3 is chosen as the upper limit of Mach number for the free
stream beyond which the effect of Mach number are to be considered.
Considering sea level standard atmospheric conditions, the maximum V will be about 100
m/s and the unit Reynolds number (for unit length) will be about 2.1x106
ft-1 (7x106 m-1).
These numbers and the appropriate characteristic length of the test article (or tunnel width)
give a good approximation to the available Reynolds number in an atmospheric wind tunnel.
Hence there has been and continues to be much attention focused on obtaining effective
results with less than full-scale articles or with various separate components tests.
An important example of the contribution of component testing is the development of airfoil

profiles for various profiles that are then incorporated into three-dimensional wing designs.
AE 2751
July-Nov. 2016
Wind Tunnel Design

Another widely used method is to test half models since many vehicles have a plane of
symmetry.
Most important of all is the careful study of aerodynamic phenomena as they are affected
by variation of Reynolds number so that useful conclusions can be obtained from tests
that do not duplicate the operating Reynolds number.
For many studies it is not necessary to produce the full-scale Reynolds number, but it
must be of a reasonable value.
For vehicles including aircrafts and racing automobiles that can operate at speeds of
Mach 0.3 or greater in the atmosphere, test articles would have to be atleast full scale to
achieve operational Reynolds numbers in an atmospheric wind tunnel.
This is impossible or very costly for many vehicles.
Much low speed testing involves aircraft takeoff and landing configurations, where the
Mach number is typically in 0.15 to 0.3 range.
Both the lift curve slope and maximum lift coefficient are affected by Mach numbers as low
as 0.2.
AE 2751
July-Nov. 2016
14
Wind Tunnel Design

This tends to require a tunnel speed approximately equal to the full scale landing speed.
In an unpressurized tunnel using air, this means that the Reynolds number ratio of model to
full scale is approximately equal to the size ratio between the scale model and the aircraft.
A primary decision is the choice of the minimum acceptable value of Reynolds number.
Because much of low-speed testing is at high-lift conditions, the effect of Reynolds number
on the airfoils at high lift must be considered.
Maximum lift and lift curve shape near stall for single-element airfoils vary considerably with
Reynolds number up to at least one million.
For multi element airfoils, this range is much greater.
The continuing need for testing facilities that allow near-full-Reynolds number of transport
aircraft high-lift systems is a primary reason for serious consideration of construction of
major new wind tunnels in the mid-1990s.
In any case, the lower boundary for testing airfoils and wings for vehicles that will operate at
higher scale values is a Reynolds number in the range of 1 to 1.5 millions based on chord.
AE 2751
July-Nov. 2016
Wind Tunnel Design

At these values of Reynolds number, the model is likely to have an extensive region of
laminar flow, and the possibility exists of poor simulation owing to separation of the nodels
laminar boundary layer.
It is assumed that laminar separations are likely to occur at full scale in the normal
operating conditions.
Therefore, flow similar to full scale is more likely to be achieved by the transition location
on the model.
If the Mach number is taken as 0.2, then the tunnel velocity is about 240 kmph (70 m/s)
For this speed, the unit Reynolds number is a little less than 1,500,000 ft-1 (450,000 m-1).
Although the minimum Reynolds number can not be rigidly defined, the above rationale
has been used to define a minimum Reynolds number of between 1,500,000 and
2,500,000 based on wing chord for low-speed tunnels to be used for aeronautical
development testing.
There are some flight vehicles that operate at lower Reynolds numbers.
There is, in fact, a whole series of airfoils for soaring gliders that are especially designed
to operate at Reynolds numbers below 1,000,000.
AE 2751
July-Nov. 2016
15
Wind Tunnel Design

There are also an increasing number of small low-speed unmanned vehicles (MAVs and
UAVs).
Operational Reynolds numbers for these aircraft are often obtainable in a medium-sized
wind tunnel.
For vehicles that operate in the atmosphere at speeds such that the Mach number is
less than 0.3, the operational Reynolds number can be duplicated in an atmospheric
wind tunnel with a scaled model.
Consider a production automobile at 60 mph (30 m/s).
The operating Reynolds number can be obtained using a three-eighth scale model with
an atmospheric tunnel test speed of 160 mph (80 m/s)
AE 2751
July-Nov. 2016
Wind Tunnel Design

Basic Decisions: Test Section Size
This is commonly the starting point in the design of a wind tunnel.
The choice will follow from considerations of the desired Reynolds number capability, the
budget for tunnel construction, and the costs for tunnel operation and tunnel users as
reflected particularly in required model characteristics.
These characteristics are discussed in the next slide.
AE 2751
July-Nov. 2016
16
Wind Tunnel Design

Airfoil models: For a mean geometric chord of 0.3 m and aspect ratio of 8-9 for an aircraft
wing, the span is about 2.5 m. The maximum span should be less than about 0.8 of the tunnel
width due to effects of tunnel walls on the flow, which leads to a width of about 3 m. For
rectangular solid-wall tunnels, wall correction factors for a small wing will be minimum for a
width-to-height ratio of about 1.5. These considerations indicate why so many tunnels have
been build in the 7x10 to 8x12 ft (2x3 to 2.4x3.6 m) size range with the maximum speed in the
range of 200-300 knots (370 to 550 kmph or 100-150 m/s). In addition to the consideration of
flow properties, a model for a tunnel of this size is large enough so that the smaller parts are
relatively easy to fabricate.
V/STOL Aircraft: For V/STOL models in a STOL descent case, the speed will be near 70
knots due to model power limits or tip Mach number on propellers and rotors. The reduction in
in test speed will require a larger model to maintain reasonable Reynolds numbers. To
minimize the wall corrections due to large downwash angles from these models, the modelspan-to-tunnel-width ratio must be smaller, typically between 0.3 and 0,5. Thus the V/STOL
tunnels built in the 1960s have test section that are 20-30 ft (6 to 9 m) wide.
AE 2751
July-Nov. 2016
Wind Tunnel Design

Automobile: A key issue for automobile tunnels is the blockage based on frontal area. The
flow around automobiles is often more characteristic of bluff bodies than of streamlined
bodies. This means there is almost a sizable region of separated flow. The wind tunnel test
sections needs to be sufficiently long so that e separated flow regions close, before end of the
test section and the entry of the diffuser. Otherwise the pressure in the separated region will not
be correct and a large influence on drag will exist. In addition, the length-to-width ratio of
automobiles is greater than for aircraft while the width-to-height ratio is much less. A wind tunnel
test section sized for automobiles is therefore typically longer than a test section sized for aircraft
and the width-to-height ratio approximates the width-to-height ratio of standard automobile.
Ideally the blockage, the ratio of model frontal area to test-section area, will be about 5% or less.
Keels, Submarines and Sails: Submarines and surface ships have large length-to-width
ratios. Their operational leeway is smaller than that for angle of attack or side slip for aircraft. A
test section chosen for these vehicles would have a length-to-width ratio of 2 or greater. Yacht
keels are sufficiently similar to aircraft wings that they fit comfortably in aircraft test sections. Test
configurations that include both keels and rudders would be better accommodated in test
sections with higher values of length-to-width ratio.
AE 2751
July-Nov. 2016
17
Wind Tunnel Design

Instructional and Other Small Tunnels: For small research tunnels and student tunnels
at universities and elsewhere, the prospect of achieving or even approaching operational
Reynolds numbers is usually beyond the available budget. In many instances the problem of
building models accurately may be a critical factor. Assuming that students can hold an airfoil
dimension of 0.25 mm and that it is sufficient to hold the model to 1-2% tolerance, the following
results can be reached. For a 12% thick airfoil with 2% tolerance, the maximum thickness
equals 1.2 cm and the chord is 10 cm. Using a mean chord of 10 cm and aspect ratio of 8. the
span is 0.8 m. AS the maximum span is 0.8 of the tunnel width, the width is 1 m. Using a widthto-height ratio of 1.5 for a minimum wall correction factor, the height is 0.7 m. The cross
sectional area is then 0.7 m2. A minimum test velocity would be about 30 m/s or a dynamic
pressure of about 50 mm of water (or 50 kgf/m2). Because of lower speeds and corresponding
dynamic pressures, instrumentation sensitivity is very low. These instruments may not be
available or very costly.
Many demonstration tunnels and calibration tunnels have test sections areas of 0.01 m2 or less.
These tunnels are very useful for observation of basic flow phenomena and calibration of
instrumentation probes. However, they are of little use for doing vehicular component studies
except in cases for which Reynolds number are quite low (ex: MAV and UAV).
AE 2751
July-Nov. 2016
Wind Tunnel Design

For a rectangular tunnel, the width determine the model size and the Reynolds number
at affixed speed. The cost of the tunnel shell and its required power tend to vary with the
square of the test section width. Since funds for a tunnel are usually fixed, the largest
tunnel that the funds will buy is generally built.
The size of smaller tunnel is frequently determined in the final analysis by the size of the
room that will house the tunnel.
Special tunnels to increase unit Reynolds number: The unit Reynolds number
can be increased by building either a pressure tunnel or a cryogenic tunnel using a cold
gas such as nitogen. Other working fluids such as Freon and sulfur hexafluoride have
beeb used to obtain desired conditions. These are special purpose tunnels, and the
need for their special capabilities must justify the cost as with any other tunnel. The time
required for model changes will be long unless special provisions are made because the
test section must be isolated before workers can enter to work on the model. Test
productivity, as well as flow characteristics, s an important characteristic of a wind tunnel.
AE 2751
July-Nov. 2016
18
Wind Tunnel Design

Open or Closed Return
Another basic design consideration is whether the tunnel will be of return or non return
(open circuit) type.
Almost all of the small research tunnels are of the non return type, usually because of
the lower construction cost.
Power consumption for such tunnels is usually not a significant factor in overall cost.
Although there exists some larger tunnels of non return design, most of the larger
tunnels are of the return types, the majority being single return.
A few of the earlier tunnels are of double return type.
However, it has been more than 70 years since the double return design has ben chosen
for a new wind tunnel.
AE 2751
July-Nov. 2016
Wind Tunnel Design

Open or Closed Test Section
Open jet tunnels will have a lower energy ratio than a closed jet wind tunnel owing to the
jet entraining stagnant air as it passes from the contraction cone exit to the collector inlet.
If the tunnel has an external balance, the balance usually has to be shielded from the air
jet and one of the boundaries tends to be closed.
Open throats do not work for an open circuit tunnel with a propeller in the diffuser unless
the test section region is enclosed in an air tight plenum or room.
Open throat tunnels offer suffer from pulsations similar to vibrations in organ pipes.
An open throat gives easy access to the model in small tunnels.
In large tunnels scaffold of some type is required to gain access to the model.
The setting up and removal of the scaffold require additional model change time.
Since the jet length is usually kept short to reduce losses, there is the possibility that highlift models may deflect the wake enough to miss the collector or that the wake of a bluff
body will interact with collector.
An open jet provides easier acess for traversing devices to move instrumentation to any
point in the flow.
AE 2751
July-Nov. 2016
19
Wind Tunnel Design

Open or Closed Test Section
In general, the advantages appear to be with the closed throat tunnel for aeronautical
testing.
However, considerations of bluff body aerodynamics and overall size requirements have
led to a number of open, partially open or slotted wall, and convertible jet tunnels built by
automobile companies and others who engage in automobile test work or V/STOL aircraft
development.
These arrangements provide greater flexibility in uses of a tunnel provided the staff are
sufficiently knowledgeable about the variety of wall effects that must be understood.
Closed thrat tunnels that are vented to the atmosphere not at the test section but at
another location will have the test section below atmospheric pressure.
Thus they can suffer from leaks either through holes cut in walls for probes, wires, pipes,
and so on, or through the struts required to mount the model.
These tunnels usually have a sealed room or plenum around the test section. Most small
open circuit tunnels are not built this way and suffer from leaks.
This makes wood an ideal material for such tunnels because it is easy to patch.
AE 2751
July-Nov. 2016
Wind Tunnel Design

General Layout: Closed Return
The general layout for closed return tunnels has reached a form generally agreed upon for
reasons of construction economy and tunnel efficiency.
Starting with the test section and going down the common configuration includes the
following elements:
a) The test section, which may be closed, open, partially open or convertible. The testsection-length-to-hydraulic-diameter ratio may typically be chosen to be 2 or more, in
contrast to the shorter test sections of earlier era tunnels.
b) A diffuser of at least three or four test section lengths. The typical equivalent cone
angle is in the range of 2-3.5O with the smaller angles being more desirable. The area
ratio is typically 2-3, again with the smaller values being more desirable.
c) First corner incorporating turning vanes.
d) Second leg that may continue the diffuser or may be constant area.
e) Safety screen to prevent parts of failed models or other unintended flying objects
reaching the fan. This screen is usually just ahead of the second-corner turning vanes.
f) Second corner incorporating turning vanes that may be essentially copies of the first
corner vanes solely to gain a small engineering and construction cost reduction.
AE 2751
July-Nov. 2016
20
Wind Tunnel Design

g) Transition from rectangular to circular cross section to take flow into the fan.
h) Fan and straightener section. Other drive devices such as ejector have also been used.
i) Return or second diffuser. This will commonly incorporate a transition back to rectangular
from the circular cross section at the fan exit. The transition will likely begun in the
straightner section. The second diffuser should follow similar design guides as the diffuser.
j) Third corner incorporating turning vanes.
k) Third leg that may be constant area.
l) Heat exchanger.
m) Fourth corner incorporating turning vanes that may be copies of the third corner vanes.
n) Wide angle diffuser with separation control screens. Typical properties are angles of about
45O and area ratios of 2-4.
o) Settling area.
p) Flow conditioners typically including flow straighteners and turbulence control screens.
q) Contaction or nozzle. Typically area ratios are in the range of 7-12, although lower and
higher values are not uncommon.
AE 2751
July-Nov. 2016
Wind Tunnel Design

A representative layout of a closed return tunnel with notation is shown below:
AE 2751
Layout of a closed single-return wind tunnel

July-Nov. 2016
21
Wind Tunnel Design

The plane of the return passage is almost always horizontal to save cost and make the
return passage easier to access.
A vertical return is justified only when space is at a premium and has only been used for
small sized tunnels.
A concept for obtaining economics in wind tunnel construction by producing a series of

various sizes for which there are many common components was developed and used.
The first- and second-corner assemblies for one size tunnel would be the same as the
third- and fourth-corner assemblies for the next smaller size tunnel.
The fan diffuser for one size tunnel would be the same as the test section diffuser for the
next smaller size tunnel.
The engineering and construction drawings would be the same for all sizes except for the
specified scale for each instance.
There are many innovative configurations to accommodate particular needs that are quite
different from
AE 2751
July-Nov. 2016
Wind Tunnel Design

General Layout: Open Circuit and Blower Tunnel
Open Circuit: This type of
tunnel is widely used for
investigations of fundamental
flow phenomena.
Blower Tunnel: This type of

tunnel also is widely used for
calibration of flow devices. A
characteristic design problem
for these facilities is choice of
details of a wide-angle diffuser
between the fan or blower and
the settling chamber before
the contraction.
AE 2751
July-Nov. 2016
22
High Speed Wind Tunnels

Definition of High Speed
When compressibility effects are pre dominant the flow is generally said to
be of high speed.
A lower limit is approximately M=0.5. Power requirements vary as cube of
velocity in the wind tunnel. This does not hold into the high speed regime
exactly.
Because of large power requirements, high speed wind tunnels are usually
of the intermittent type.
This material is taken from the course notes of Experimental Aero (Gas) dynamics
by Prof. Job Kurian of Department of Aerospace Engineering, IIT Madras
AE 2751
July-Nov. 2016

Types of High Speed Wind Tunnels
1 Continuous (for all speed ranges)
2 Intermittent
2.1 Blowdown: M > 0.5 < 5.0
2.2 Indraft
2.3 Intermittent pressure vacuum tunnel for M>5
AE 2751
July-Nov. 2016

Pressure Driven Blow Down Wind Tunnel
.
AE 2751
July-Nov. 2016

Indraft Type Wind Tunnel
AE 2751
July-Nov. 2016

Comparison between Indraft and Pressure Driven Wind Tunnels
Indraft wind tunnels
Pressure driven wind tunnels
Stagnation temperature at supply condition Reynolds number can be varied at a

particular Mach no.
is constant during a run.
So also is total pressure. No fluctuations as
those generated by a pressure regulator.
No possible contamination such as that due Cost is much less than of an indraft tunnel.
to oil.
Vacuum is safer to handle than pressure.
Pressure regulators are not needed.
The wind tunnels described above can be converted as continuous tunnels.
The comparison between blow down and continuous wind tunnels are given in next slide.
AE 2751
July-Nov. 2016

Comparison between Intermittent and Continuous Wind Tunnels
Intermittent (blow down) wind tunnels
Continuous wind tunnels
Simple to design and less costly
More in control of conditions and return to

a given test condition with more accuracy.
A single drive may run several tunnels
Check points are easily obtained

No panic of rapid testing
Test conditions can be held constant for a
longer time.
Model testing is more convenient

Extra power is available to start
Failure of model will not result in tunnel damage
AE 2751
July-Nov. 2016

Supersonic Wind Tunnels: Introduction
.The nozzle regulates the speed of air entering the test section of the wind tunnel so that
the desired Mach number is established. Mach number is uniquely determined by the
area ratio of the nozzle. A well designed nozzle makes the flow parameters uniform
across the cross section. The design of a suitably shaped nozzle contour to obtain the
desired uniform flow at the nozzle exit is based on the method of characteristics.
AE 2751
July-Nov. 2016

Supersonic Wind Tunnels: Test Section Parameters
Dynamic pressure: The local dynamic pressure 1/2v2 can be related to the local
Mach number and static pressure.
Mass flow rate: The mass flow rate is given as m=vA=*a*A*

where * represents choked or sonic conditions (M=1)
AE 2751
July-Nov. 2016

.
AE 2751
July-Nov. 2016

Writing in terms of stagnation conditions
AE 2751
July-Nov. 2016

. flow density is expressed as a function of stagnation conditions and the area ratio.
Mass
For isentropic flow, A* is a constant and A/A* is a unique function of local Mach number.
AE 2751
July-Nov. 2016

.
Test section velocity;
Maximum velocity;
or
AE 2751
v=aM=
is maximum when cpT=0
cp-cv=R
cp/cv=g
cp-cp/g=cp(1-1/g)=(cpg-cp)/g=cp(g-1)/g=R
cp=gR/(g-1)
July-Nov. 2016

The test section flow velocity v for a given stagnation temperature TO approaches the
maximum value vmax at relatively low supersonic Mach numbers.
For example, in the case of TO = 300 K R = 287J/kgK and a test section Mach number
of 5.0, the ratio of v/vmax can be calculated to see that it is equal to 0.913.
This means at ordinary stagnation temperatures, the velocity in the test section reaches
91% of the maximum possible velocity corresponding to the total energy of the fluid.
The stagnation temperature T0 rather than the Mach number which is important to
attain high velocities.
AE 2751
July-Nov. 2016

Free stream Reynolds Number (Re):
Re=vL/
Experimental observation is that is independent of pressure in the range of 0.001 to

20 atmospheres.
AE 2751
July-Nov. 2016

Free stream Reynolds Number (Re):
Re=vL/
If this relation is assumed then the free stream Re can be expressed as a function of
M1, the test section Mach number and of the stagnation parameters, Reynolds number
per unit length Re/L=v/ are expressed in stagnation quantities as given below.
M=v/a
AE 2751
v=aM=a
July-Nov. 2016

.
n= 0.768 for air

Simplifying
Both ao and o are functions of stagnation temperature. Both increase with
temperature.
Hence, appreciable changes in free stream Re/unit length for a given M can be
obtained only by varying stagnation density.
AE 2751
July-Nov. 2016

Components of a Supersonic Wind Tunnel
Air storage tanks: Size of the storage will be dependent on the mass flows required
.
and the frequency of runs.
Pressure storage tanks are available on the shelf basis They are mounted horizontally
or vertically. Tanks are painted black to absorb heat.
They are provided with safety disk or pressure relief valve.
As air is drawn from the storage, polytropic expansion takes place within the tank.
This results in drop of reservoir temperature which is very bothersome.
Fall of stagnation temperature causes resultant change in the stream temperature for a
given Mach number. Change in temperature results in the change of viscosity which in
turn affects the boundary layer thickness. Changes in Reynolds number and Mach
number during a run are thus consequential to the fall in reservoir temperature.
To maintain constancy of stagnation temperature, it is a practice to stack the reservoir
volume with empty metallic cans. They serve as heat storing matrix during compression
and release heat during the expansion process. Another way to maintain the constant
stagnation temperature is by providing heater units in the reservoir.
AE 2751
July-Nov. 2016

Settling chamber /wide angle diffusers: Wide angle diffusers lead the flow to the
.
settling chamber. Arrangements for leading the flow to the settling chamber may be by one
of the methods shown below:
a, b: Wide angle diffusers
c: Reverse entry into the settling chamber
Uniformity of flow in the test section is improved if a large area ratio contraction is provided.
AE 2751
July-Nov. 2016

Convergent-divergent (CD) nozzle: The CD forms the heart of the supersonic
.
wind tunnel .For generating supersonic flow in the test section, it is essential that there
is a c-d nozzle in the tunnel circuit before the test section.
The area ratio of the CD nozzle (Aexit/Athroat) uniquely decides the Mach number
Convergent-divergent nozzle and the pressure profile
AE 2751
July-Nov. 2016

Convergent-divergent
(CD) nozzle: When the tunnel
.
This is called the choked condition of the nozzle.
Isentropic flow, g=1.4
operation starts, the flow is initiated in the nozzle as

subsonic and reaches the sonic Mach number at the throat
when sufficient mass flow is allowed.
Under this condition, maximum mass flow rate for the given
stagnation conditions takes place through the nozzle.
The ratio between the upstream stagnation pressure (PO)
and the downstream back pressure (Pb) corresponding to
the first time choking is called the first critical pressure ratio
of the nozzle.
At this pressure ratio, the flow in the divergent part of the
nozzle is subsonic.
The exit Mach number will be the subsonic value
corresponding to the A/A* shown in the figure.
AE 2751
A
A*
1
0
0
M<1
M>1
Area ratio, A/A* vs. Mach number
July-Nov. 2016
10

Convergent-divergent
(CD) nozzle: In this context A is the exit area and A* the
.
throat area of the choked nozzle.
As the value of PO/Pb is progressively increased, the flow in the divergent part of the
nozzle accelerates to be supersonic but shocks are formed in the divergent part until a
pressure ratio corresponding to the supersonic Mach number of the nozzle is reached.
The pressure ratio corresponding to this Mach number is the third critical pressure ratio
of the nozzle.
Between the first and third critical pressure ratios shocks of varying strengths take place
in the nozzle and outside of it as there is no other isentropic solution between the 1st
and 3rd critical pressure ratios.
The pressure ratio corresponding to the occurrence of a shock at the exit plane of the
nozzle is the second critical pressure ratio of the nozzle.
AE 2751
July-Nov. 2016

. Convergent-divergent (CD) nozzle: All the shocks generated at the different
pressure ratios inside the nozzle will make the post shock Mach number subsonic and
the subsonic nozzle exit pressure will be made equal to the ambient pressure in the
remaining part of the diffusing divergent channel.
Between the second and third critical pressure ratios, oblique shocks of varying
strengths depending on the pressure ratio will be formed emanating from the nozzle lip.
The physical purpose of these oblique shocks is equalization of pressures between the
exit plane and the ambient. If the pressure ratio is increased beyond that corresponding
to the third critical pressure ratio, expansion fans will be formed at the lip of the nozzle.
AE 2751
July-Nov. 2016
11

Diffuser-the
. necessity of providing a diffuser
Take the case of a free jet facility shown in making

use of a CD nozzle of Mach number 3, exiting to the
ambient conditions at one atmosphere pressure.
In order to avoid shocks and expansion waves at the

exit of the nozzle, pe must be pb.
(po/pe)M=3=36.7
Pressure ratio required will be 36.7 for a wave free
exit flow from the nozzle
Nozzle of a free jet facility
AE 2751
July-Nov. 2016

A constant area section is added to the nozzle exit.
.
The duct similar to the test section of a wind tunnel
attached to the nozzle exhausts to atmosphere.
pe corresponds to static pressure at the exit plane
of the nozzle before the shock.
Static pressure after the shock (p2) is equal to
ambient pressure.
po/p2=(po/pe) (pe/p2)=36.77/10.33=3.55
In the equation above, pe/p2 represents the shock
pressure ratio at M=3.0.
AE 2751
Free jet nozzle with a test section
July-Nov. 2016
12

In the third case, a divergent channel is provided after the constant area
.duct and the shock stands at the end of the constant area duct.
Free jet facility with test section and a diffuser
AE 2751
July-Nov. 2016

.W hen M << 1, p=p
po/p=(po/pe) (pe/p2) (p2/po2=p)M=0.475=36.77/10.33X0.856=3.04

The three cases described above make it clear that provision of a diffuser of
suitable design is required for reducing the pressure ratio required for the operation
of the wind tunnel.
It will be shown that the power required to run the wind tunnel increases with the
pressure ratio.
In supersonic wind tunnels, most commonly used diffuser is of convergent divergent
type (also called the second throat diffuser).
AE 2751
July-Nov. 2016
13

Power required for the operation of supersonic wind tunnel
.
Free jet type wind tunnel with an attached test section
AE 2751
July-Nov. 2016

.Refer to the free jet type wind tunnel shown. Let the supersonic tunnel is specified by
the Mach number (M) in the test section and test section (A) area. The throat area is
specified as ANT.
The flow parameters in the test section are denoted as p, T, A etc. Once M is specified,
the area ratio AT/ANT and ratios of pressure and temperature p/p01 and T/T01 are all
known from the isentropic equations. If the compressor is idealized (isentropic) the
suction and reservoir conditions are related.
ps and Ts represent pressure and temperature at the compressor suction.

If the nozzle is choked, the mass flow rate can be written as,
AE 2751
July-Nov. 2016
14

.The power required to operate the compressor may be found as follows:
If the compressor is isentropic, the work per unit time may be found from the enthalpy
difference across the compressor.
AE 2751
July-Nov. 2016

.
The pressure ratio, po1/ps is known as the operating pressure ratio and is denoted by rp.
The very large power required for the operation of a supersonic wind tunnel is attributed
to the large operating pressure ratio.
If the wind tunnel is equipped with a suitably designed diffuser and a closed circuit
arrangement as shown later, the stagnation pressure of the diffused high velocity air can
be made use of by the compressor and the effective pressure ratio can be reduced.
AE 2751
July-Nov. 2016
15

Closed circuit supersonic wind tunnel
.
AE 2751
July-Nov. 2016

.Assuming a frictionless operation, the shock may assume any section in the constant
area test section.
But, the effect of friction is to make the shock unstable in the constant area duct.
The shock that is generated during starting of the tunnel does not stay at the nozzle exit
(entry to test section) but is moved downstream by the effect of friction.
A convergent-divergent (CD) diffuser is provided.
Only if the flow is isentropic, the stagnation pressure regained in the receiver following
the diffuser (p02) is equal to that of the flow entering the nozzle (p01).
In that case, p02 = p01. Then, rp = 1, so that compressor does no work.
AE 2751
July-Nov. 2016
16

.In practical cases, because of entropy changes p02 < p01 and rp > 1.
If the entropy change is confined to the region between the two throats, the
diffuser throat area ADT must be larger than the nozzle throat area ANT.
Diffuser throat area must be large enough to accommodate the stagnation pressure
loss of the strongest shock.
A cooler is included prior to the compressor because compressor work is proportional to
the intake temperature.
AE 2751
July-Nov. 2016

The practical operation of the closed circuit wind tunnel may be explained as follows:
As the tunnel is started, flow through it begins as subsonic and as the pressure ratio is
increased the nozzle is choked.
Further increase in the pressure ratio causes shock to be formed in the divergent section.
At a pressure ratio corresponding to second critical pressure ratio, shock is
formed at the exit plane of the nozzle which is same as the entry section to the test section.
The formation of shocks during the starting process gives rise to fall in stagnation pressure.
The total pressure after the shock is designated as p02.
This necessitates that the diffuser throat is designed larger as decided by the ratio p02/p01.
The ratio of diffuser throat area to the nozzle throat is in the inverse ratio of total pressures
given above.
AE 2751
July-Nov. 2016
17

The ratio of areas for different test section
Mach numbers calculated based on
normal shock losses is given in the figure.
This makes sure that the starting shock
passes through the diffuser throat.
The diffuser throat area calculated as
above does not take in to account the non
isentropy of frictional flows and only the
shock losses are considered.
1.0
0.8
ADT
ANT
0.6
Mach Number, M
Ratio of diffuser throat and nozzle throat

for different test section Mach numbers
AE 2751
July-Nov. 2016

Shock movement from nozzle exit to diffuser

Assuming a frictionless operation, the shock may assume any section in the constant
area test section.
But, the effect of friction is to make the shock unstable in the constant area duct.
The shock that is generated during starting of the tunnel does not stay at the nozzle exit
(entry to test section) but is moved downstream by the effect of friction.
AE 2751
July-Nov. 2016
18

The starting pressure ratio minimum required to cater to the shock at the test section Mach
number is corresponding to that for locating the shock at the nozzle exit. (worst shock).
rps=(po1/po2)worst hock
As the starting pressure ratio is maintained, the starting shock which moves down stream can
be stable only at an area equal to that of the test section.
In the convergent part of the diffuser the Mach number will be less than that in the test section.
With the value of starting pressure ratio which produced the shock at the test section Mach
number being maintained, shock losses at that Mach number is being catered to.
Hence the starting shock stabilizes only in the diverging part of the diffuser at a section where
there is equal area and Mach number as the test section.
The starting shock crossing the diffuser throat and remaining in its divergent part is
called the swallowing of the starting shock.
AE 2751
July-Nov. 2016

AE 2751
rting
30
Sta
20
Op
er
at
in
g
DT
Therefore the higher pressure ratio is

required only during the starting when a
shock at the test section Mach number is
necessarily to be catered to and thereafter
the pressure ratio can be that corresponding
to shock at the diffuser throat.
40
Pressure Ratio, rp
It has to be remembered that as the diffuser

throat is larger than the nozzle throat, the
Mach number there will be more than one
but less than that in the test section.
After the brief duration of starting, the
pressure ratio may be decreased and the
shock may be brought to the diffuser throat.
rpo=(po1/po2)shock at M
10
0
1
Throat Mach Number, MT

Pressure ratios for starting and operating
the wind tunnel of different Mach numbers
July-Nov. 2016
19

In summary, the pressure ratio for starting the wind tunnel is corresponding to
the normal shock losses at the test section Mach number and that for operation
is that corresponding to normal shock losses at the diffuser throat.
Hence the power required can be considerably reduced by incorporating the well
designed diffuser and by judicious control of the two pressure ratios during
starting and operating of the wind tunnel.
AE 2751
July-Nov. 2016

Actual flow in the supersonic wind tunnel
The boundary layer thickness and the total loss of momentum increase with increasing
distance from the 1st throat.
The growth of boundary layer thickness with distance from first throat is predictable and
can be accounted for in the nozzle design.
In the steady state operation, viscous effects between the throat and test section are not
of much importance.
During the transient process in which tunnel is started, viscous effects are much
important.
So, important are these effects that pressure ratios required to start high Mach number
tunnels are atleast 100% greater than the normal shock pressure ratio, po1/po2.
i.e. viscous losses are almost equal to normal shock losses.
Boundary layer is stable when the pressure is decreasing in the direction of its growth.
When the pressure is increasing in the direction of flow, it has a tendency to separate.
AE 2751
July-Nov. 2016
20

As normal shock passes through the nozzle, it imposes a severe unfavorable
pressure gradient which can cause separation.
If boundary layer separates, it disturbs the flow over a large portion of nozzle.
If boundary layer does not separate the high pressure gain in the downstream of
shock will tend to flow to low pressure boundary layer and the flow in the duct will be
altered over a significant length of the nozzle.
In the diffuser viscous effects are predominant during starting and steady state
operation of the wind tunnel.
Unfavorable pressure gradient exists always. An oblique shock from the convergence
creates additional pressure gradients when they strike the opposite wall.
AE 2751
July-Nov. 2016

As normal shock passes through the nozzle, it imposes a severe unfavorable pressure
gradient which can cause separation. If boundary layer separates, it disturbs the flow
over a large portion of nozzle. If boundary layer. does not separate the high pressure
gain in the downstream of shock will tend to flow to low pressure boundary layer. and
the flow in the duct will be altered over a significant length of the nozzle. In the diffuser
viscous effects are predominant during starting and steady state operation of the wind
tunnel. Unfavorable pressure gradient exists always. An oblique shock from the
convergence creates additional pressure gradients when they strike the opposite wall.
AE 2751
July-Nov. 2016
21

Starting a tunnel with a model in the test section
It is explained earlier that
which implies that losses in total head resulting from shocks necessitate a larger
diffuser throat.
Hence, losses due to shocks on the model must also be provided for. So, for starting
a tunnel with a model, a second throat larger than that for a clear tunnel is needed.
AE 2751
July-Nov. 2016

Sizing of the wind tunnel model
The theoretical unobstructed cross section area of the test section at the model required
for starting is the same as the second throat area.
In choosing the dimensions of the model, reflection of shocks should be also considered.
The oblique shocks formed as shown in figure at the leading edge of the model get
reflected from the wind tunnel wall.
It has to be remembered that shock reflection is not specular which means that the angle
of incidence of the shock at the test section wall is not same as that for reflection.
The chord length of the model is so
chosen that the reflected shocks do
not interfere with the model.
Typical shock wave pattern from a model
AE 2751
July-Nov. 2016
22

Problems in the operation of supersonic wind tunnels
Condensation: The amount of moisture that can be held by a unit volume of air
increases with increasing temperature. When the air isentropically expands to higher
Mach numbers in the test section, the temperature falls.
It may become super cooled. Moisture will then condense.
Factors affecting condensation:

a) Amount of moisture in the stream b) Static temperature of the stream
c) Static pressure of the stream
d) Time during which the stream is at low
temperature
Effects of condensation: Condensation results in changes of local Mach number and

other flow properties due to latent heat addition. The extent of changes depends on how
much heat is released through condensation and may be evaluated using the two
equations given below:
AE 2751
July-Nov. 2016

where dQ = heat added through condensation

H = enthalpy per unit mass
A = duct area
When M > 1, Mach number decreases and pressure increases.
When M < 1, Mach number increases and pressure decreases.
Drying the working fluid is the best way to avoid condensation.
Increasing the temperature by providing stagnation heaters is another solution
AE 2751
July-Nov. 2016
23

condensation, the components of air liquefy
when proper temperature and pressure
conditions are met.
Liquefaction troubles might start around M=4
if high pressure air is expanded from room
temperature.
The figure shows the stagnation temperature
required to avoid liquefaction at different Mach
numbers.
It can be seen that corresponding to a Mach
number above 12 the temperature
required will be about 2000K
AE 2751
2000
Stagnation Temperature
to avoid Liquefication (K)
Liquefaction: In a manner parallel to
1500
1000
500
10
Mach Number, M
12
14
Stagnation temperature to avoid

liquefaction at different Mach numbers
July-Nov. 2016
24

Introduction
After a tunnel is constructed, the next step is to determine its flow characteristics and, to change
any that are not satisfactory for the purposes intended.
The low
low-speed
speed "steady"
steady airstream is usually considered to be defined when we know its
distribution of temperature T, pressure p, dynamic pressure q, and "turbulence" Tu.
There are several assumptions embedded in this statement. We admit that the airstream is
expected to be turbulent and therefore not strictly steady or time invariant. The time variability of
the airstream is to be characterized by its level of turbulence, which is commonly defined to be
the rms of the variation of the longitudinal component from the mean value of air speed.
The pressure and dynamic pressure are the mean values for these quantities averaged over a
time interval chosen as necessary to achieve the required precision of the mean. The stream
temperature is similarly considered. Using the equation of state, we can then compute the
density, and using the definitions of total pressure and total temperature, we can compute these
quantities and flow speed as well.
AE 2751
July-Nov. 2016

Introduction
In the case of wind tunnels to be used for wind engineering, the temporal and spatial
variations must be considered in considerably more detail and will include spectra and
integral scales of the turbulent flow structures as well as a profile to match appropriate
types of planetary boundary layers
layers.
We may also compute the Reynolds number for a particular model based on its chosen
characteristic length. Much of our interest is centered on determining pressure, which
can be measured most simply by use of liquid manometers or more commonly by using
electronic data systems, including various instrumentation elements and computer
systems for data manipulation, storage, and presentation. For our subsonic testing, the
precision with which measurement of stream temperature are made are usually less
demanding than the typical pressure measurements.
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
Setting the speed of a wind tunnel appears straightforward as we apply our basic understanding
of fluid dynamics. However, it turns our that this basic question absorbs a considerable amount
of attention even for experienced aerodynamicists.
aerodynamicists When there is no model in the rest section
section, a
measuring device, most commonly a Pitot-static tube, can be put there to determine the air
speed. One cannot, however, insert a Pitot-static rube or other measuring device in the test
section to measure dynamic pressure or speed along with an object under test because the test
object will cause changes in the flow. These changes are referred to as "induced flow."
Consider Fig. 1, which gives a schematic indication of the settling chamber, contraction and test
section of a typical wind tunnel. The tunnel speed is usually determined by measuring either
static or total pressure in the settling chamber ahead of the contraction cone,
cone as indicated by
station L, and a static pressure ahead of the test section, as indicated by station S. If
honeycomb or screens are used in the settling chamber, station L will be downstream of these
devices. Using the subscripts L for the bellmouth or settling chamber and S for the region before
the test section, Bernoulli's equation between the two stations will be
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
Fig. 1 Typical measurement stations for a

wind tunnel q system
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
A better approach would be to survey a volume of the test section that is occupied by the
model. If the Pitot-static tube used has a hemispherical nose, it can be equipped with two
yaw heads at 90
90 to each other, and the distribution of the upflow and cross-flow
cross flow can be
obtained simultaneously.
The survey of the lest section can be done with a simple Pitot-static tube assuming there is
a method to position it throughout the test section or a rake of Pitot statics can be used. lf
there is no method of remotely positioning the Pitot-static tube, the survey of the test section
becomes a very tedious operation. In any event, this is a critical operation and the effects of
the manner of holding and supporting any selected instrument must be carefully evaluated.
If a volume of the test section is surveyed and there are large variations in the dynamic
pressure it may be desirable to use a weighted average of the measured dynamic
pressure,
pressure. When the dynamic pressure calibration is completed, there is a relation
established between the indicated dynamic pressure (PL - PS) and the dynamic pressure q,
in the clean test section. There usually are a series of these calibrations for different testsection configurations. If the tunnel has an external balance, calibrations are often made
with and without the balance struts and their fairings. If a ground plane is used often,
calibrations are made for various heights of the ground plane.
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
It is desirable to obtain for each test-section configuration the distribution of the total
pressure, static pressure, upflow, and cross-flow throughout the region occupied by
the model.
model When the tunnel has an air exchanger or heat exchanger
exchanger, the temperature
distribution should also be measured.
The pressure sources for the tunnel speed control should be, ideally, either a ring
around a tunnel station or at least a portion of a ring on the two side walls or the
ceiling and floor. The possibility of the model pressure field directly affecting the
nearer of the pressure sources at station S should always be kept in mind and
evaluated if a larger model than normal or an upstream mounting location is to be
considered. The static source should consist of either a series of static sources or a
sealed tube flush with the surface with many holes evenly spaced along its length.
The tubes or multiple static sources of the ring should be manifolded together to yield
an average static pressure at the station. If total pressure is used in the contraction
cone. It is desirable to have multiple sources also.
AE 2751
July-Nov. 2016

Test Section Speed
Speed Setting
The two pressure sources in the simplest case can be connected to a U-tube
manometer that is used for setting the tunnel speed. It is more common to use a highquality
q
y differential p
pressure transducer. Then the electrical output
p can be transmitted
to any location convenient for the tunnel operator and to the data system.
Extreme care must be taken to make sure that there are no leak in the tunnel dynamic
pressure system. In large tunnels, the leak checking can be a time-consuming
process. It is highly desirable and strongly suggested that the system should not be
disturbed once it has been leak checked. It is often useful to record the pressures at
each of the two tunnel stations as separate readings relative to atmospheric ambient
pressure. Additional sources for this use should be provided rather than tapping into
the tunnel dynamic pressure system.
It is also desirable to have a simple method of periodically checking the system for
leaks. One way to do this is to use a calibration wing that can be installed easily. The
wing can be run through a pitch series at several dynamic pressures. If the slope of
the lift curve does not change, there is no change in the dynamic pressure calibration.
This has the further advantage of also checking the tunnel upfIow. If the drag polar
does not rotate, the upflow has not changed.
AE 2751
July-Nov. 2016

Horizontal Buoyancy
As the air proceeds along the test section, the boundary layer thickens. This action reduces the
effective area of the jet from that of the geometric dimensions and causes an increase in the flow
speed outside the wall boundary layers. If the geometric area is constant, the speed increase in
produces a drop
p in local static p
pressure, tending
g to draw the model downstream. This added
turn p
drag is commonly called "horizontal buoyancy" as its action is analogous to the buoyancy due to
the vertical pressure gradient in the atmosphere and the ocean. If the cross-sectional area of the
jet is increased enough to allow for the thickening boundary layer, a constant value of the static
pressure may be maintained throughout the test section. No exact design method is available
that ensures the development of a constant static pressure. For a first approximation the walls of
a closed jet should diverge about to each; finer adjustments may be necessary after the tunnel is
built and the longitudinal static pressure is measured. Some tunnels whose test sections have
corner fillets have these fillets altered until a constant static pressure is obtained. The
advantages
d
t
off such
h a flflow are enough
h tto jjustify
tif a moderate
d t amountt off work
k iin obtaining
bt i i itit.
The amount of "horizontal buoyancy" is usually insignificant for wings, but for fuselages and
nacelles it is larger and becomes important. For large torpedo like bodies, it can be of the order
of the minimum drag in some tunnels.
AE 2751
July-Nov. 2016

Longitudinal Pressure Gradient
The static pressure gradient along the test section must be known in order to make necessary
buoyancy corrections. It may be obtained by reading the local static pressure with a Pitot-static
tube that is progressively moved from entrance cone to exit cone. Care must be taken that the
Pitot tube is aligned directly into the wind and that no extraneous static pressure is created by
th bracket
the
b k t holding
h ldi th
the Pit
Pitott ttube.
b Thi
This llastt proviso
i iis nott possible
ibl tto satisfy
ti f exactly.
tl
Perhaps a more convenient method is to use a long static tube, as discussed in the next
paragraph. It is not desirable to measure the static pressure along the walls of the tunnel. It
turns out that small variations in wall geometry associated with joints, mounting holes, and the
like cause local variations in the static pressure that are typically too large to ignore and may
change over time as various mounting installations are put in and taken out of a facility.
Measuring wall pressures is a good adjunct to calibration and has become a common practice
to obtain input for boundary corrections, but it is not a good idea to rely on wall pressure
measurements for the baseline characterization.
characterization
Long Static Tube: This is a long tube that extends through the test section. The tube is often
suspended on a series of wires that are used to align and tension the tube. The tube is
equipped with a number of static pressure rings. These rings have four or more static ports
around the circumference that are manifolded together. If the flow angularity is small, then aft
of the nose the flow will be practically parallel to the tube. This then yields a static pressure
distribution along the length of the tube.
AE 2751
July-Nov. 2016

Dynamic Pressure Variation
The dynamic pressure is usually measured throughout selected regions of the test section
by means of a Pitot-static tube. The density is calculated from a barometric pressure
measurement. A test-section static pressure measurement relative to the atmosphere, and
a test-section temperature determination along with the equation of state.
The local velocities are then obtained from V=(2q/)0.5(7).

The velocities as calculated from the dynamic pressures
or the pressures themselves are plotted, and the points
are connected by "contour" lines of equal values. The
variation of q in the working range of the jet should be
less than 0.50% from the mean, which corresponds to a
0.25% variation in velocity. Some tunnels have been
built to tighter specifications. But it is almost inevitable
that custom tweaking of aspects of the tunnel circuit will
be required to do better.
A plot of the dynamic pressure distribution in a
rectangular test section is shown in Fig. 2. Of interest is
the asymmetry that is usually found, and the maximum
variation in this example is well above satisfactory
limits. The survey should have been carried to the walls. Fig. 2 Distribution of dynamic pressure
in the test section
AE 2751
July-Nov. 2016

Dynamic Pressure Variation
The correction of an excessive velocity variation is not as serious a problem as the correction
of excessive angular variation. There are more method of attack as well as less probability
that the variation will change with tunnel speed. There are several minor adjustments that
may be expected to improve a less than satisfactory speed distribution.
distribution There may be local
flow separations that must be found and corrected or improperly set turning vanes.
If the velocity variation is annular, the source may
be the propeller load distribution. Such local
problems in identifiable parts of the tunnel should
be corrected. Finally screens may be added in the
largest section of the tunnel with spatial mesh
densities
de
s t es varied
a ed so tthat
at tthey
ey a
are
e more
o e de
dense
se in tthe
e
sections that correspond to high-velocity regions in
the jet. The improvement in velocity distribution by
such screens is shown in Fig. 3. The loss in energy
ratio they cause is quite small and is far outweighed
Fig. 3 Effect of screens on velocity
by the improvement in testing conditions.
distribution in the test section
AE 2751
July-Nov. 2016

Flow Angularities
The variation of flow angle in the jet can be measured by many devices. Regardless of
the device used, it is desirable to map the upflow and cross-flow in a series of transverse
planes along the longitudinal axis in the region occupied by the model over the range of
intended dynamic pressures. Often, when the upflow and crossflow are plotted as flow
direction vectors, regions of vortex like flow can be seen in the test section. Such flow is
often the result of poor velocity distribution in the return duct before the third corner or the
result of improperly set fourth-corner turning vanes.
A variation of upflow across the span of a wing results in an effective aerodynamic twist. A
cross-flow gradient across the test section in the region of the vertical tail will change the
slope of the yawing moment versus side slip or yaw angle
angle. Thus it is desirable to have the
upflow and cross-flow constant across the tunnel. This is difficult to achieve. It would be
desirable to have the variation less than + 0.10O, but it is often necessary to accept the
best values that can be achieved. The maximum variation should be held to + 0.10O.
AE 2751
July-Nov. 2016

Turbulence Measurements
Variations between results of tests made in different wind runnels at the same Reynolds
number and between tests made in wind tunnels and in flight have indicated that some
correction was needed for the effect of turbulence that exists in wind tunnels. It has
b
been
argued
d th
thatt this
thi tturbulence
b l
causes flflow patterns
tt
iin th
the tunnel
t
l to
t be
b similar
i il tto th
the
flow pattern in free air at a higher Reynolds number. Hence the tunnel test Reynold
number could be said to be a higher "effective Reynolds number.
The physics of turbulent flow is far too complex to be captured by this simple concept.
There are some phenomena for which it "works" to some extent and others for which it
does not. This concept is dependent on the long-known fact that spheres (and circular
cylinders)
li d ) h
have quite
it well-defined
ll d fi d critical
iti l Reynolds
R
ld numbers
b
near which
hi h th
the d
drag
coefficient drops rather precipitously as the Reynolds number increases.
AE 2751
July-Nov. 2016

Turbulence Measurements
It has been experimentally verified that the Reynolds number at which the drag
coefficient of a sphere decreases rapidly depends strongly on the degree of turbulence
in the wind tunnel. The Reynolds number at which the reduction occurs decreases with
increasing tunnel turbulence
turbulence. There is also a strong effect of surface roughness.
roughness We are
considering only aerodynamically smooth spheres. The decrease in drag coefficient
with Reynolds number can be understood as the result of increasing the Reynolds
number, producing earlier boundary layer transition from the laminar to turbulent state,
which in turns leads to a downstream shift in the separation point and corresponding
higher base pressure, a smaller wake, and less drag. Early night measurements on
spheres found that in the free atmosphere the critical Reynolds number for a sphere is
3.85X105. This value is larger
g than is achieved in wind runnels, although
g many
y lowturbulence tunnels approach this value. In the atmosphere the turbulent eddies are so
large relative to the sphere that they do not affect the thin boundary layer of the sphere.
AE 2751
July-Nov. 2016

Associated Instrumentation
AE 2751
Fig. 4 Turbulence sphere

Drag Coeffic
cient of Sphere
Before the now common use of hot-wire

anemometry, a turbulence sphere was the
primary way to measure the relative
t b l
turbulence
off a wind
i d ttunnel.l It remains
i a very
useful and easy way to characterize the
turbulent environment in a tunnel and to
check if there is an indicated effect following
changes in the tunnel configuration or special
installations that might affect the flow quality.
The configuration of a turbulence sphere is
sho n in Fig
shown
Fig. 4
4. The critical Re
Reynolds
nolds
number for the sphere can be measured in
two ways. One method is to plot the
measured CD based on cross-sectional area
versus Reynolds number, as shown in Fig. 5.
0.5
0.4
0.3
0.2
Critical Reynolds
Number=336,000
0.1 5
2x10
3x10
Reynolds Number of Sphere
4x10
Fig. 5 Variation of drag coefficient for a turbulence

sphere as a function of Reynolds number
July-Nov. 2016

Before the now common use of hot-wire anemometry, a turbulence sphere was the
primary way to measure the relative turbulence of a wind tunnel. It remains a very
useful and easy way to characterize the turbulent environment in a tunnel and to check
if th
there is
i an indicated
i di t d effect
ff t following
f ll i changes
h
iin th
the ttunnell configuration
fi
ti or special
i l
installations that might affect the flow quality. The configuration of a turbulence sphere is
shown in Fig. 4. The critical Reynolds number for the sphere can be measured in two
ways. One method is to plot the measured CD based on cross-sectional area versus
Reynolds number, as shown in Fig. 5. From the plot the Reynolds number in the tunnel
for CD = 0.3 is read. The second method is to take the average of the four pressures on
the aft surface of the sphere and subtract this value from the stagnation value at the
leading edge of the sphere,
sphere yielding
ielding P.
P A plot of P/q versus
ers s Re
Reynolds
nolds n
number
mber is
prepared for the sphere and the Reynolds number determined for P/q = 1.22, as
indicated in Fig. 6. The pressure method has certain advantages. It needs no drag
balance with the associated balance calibration and no evaluation of support tares for
the portion of the support sting exposed to the airstream. Also, the sphere support sting
can be stiffer as no deflection is needed by the drag balance.
AE 2751
July-Nov. 2016

1.4
Prressure Loss Coefficient, P/q
From the plot the Reynolds number in the

tunnel for CD = 0.3 is read. The second method
is to take the average of the four pressures on
the aft surface of the sphere and subtract this
value from the stagnation value at the leading
edge of the sphere, yielding P.
P/q=1.22
12
1.2
A plot of P/q versus Reynolds number is

prepared for the sphere and the Reynolds 1.0
number determined for P/q=1.22, as indicated
Critical Reynolds
in Fig. 6. The pressure method has certain
Number=299,000
advantages. It needs no drag balance with the 0.8 5
5
5
5
5
1 5x10
2 0x10
2.0x10
2 5x10
2.5x10
3 0x10
3.0x10
3 5x10
3.5x10
associated
i t d balance
b l
calibration
lib ti
and
d no 1.5x10
Reynolds Number, Re
evaluation of support tares for the portion of the
support sting exposed to the airstream. Also, Fig. 6 Variation of pressure coefficient
for a turbulence sphere as a
the sphere support sting can be stiffer as no
function of Reynolds number
deflection is needed by the drag balance.
AE 2751
July-Nov. 2016

The critical Reynolds number as defined by either force or pressure measurements
is then used to define a turbulence factor for the tunnel by comparing the tunnel's
critical Reynolds number to the atmospheric free-air Reynolds number:
TF 385 000/R tunnel
TF=385,000/Re
Then the effective test Reynolds number is defined by Reeffective=TFXRNtest
The use of a turbulence sphere yields what may be thought of as an average value
of tunnel turbulence. It does not give any information on the magnitude of
turbulence in either the axial or lateral direction. The use of a turbulence sphere
may, however, prove to be a simple method of monitoring any change in tunnel
turbulence. Its use requires
q
no p
prior calibrations and the installation and running
g in
a runnel can be designed to be simple and quick.
The relation between the critical Reynolds number of a sphere and turbulence
intensity as measured by a hot wire is shown in Fig. 7.
AE 2751
July-Nov. 2016

Turb
bulence Intensity, TI (%)
Turbulence spheres can be made from cue, duck, and bowling balls. Several sizes are
needed to enable the turbulence factor to be measured over a range of tunnel air speeds.
2.4
A brief examination of Fig. 7 might lead to the conclusion
th t the
that
th higher
hi h th
the tturbulence,
b l
th
the b
better
tt th
the ttunnel,l as th
the
2.0
effective Reynolds number of the test would be higher,
This correction is not exact and if the tunnel has excessive 1.6
turbulence, the model may have a premature transition
from laminar to turbulent flow, which can be critical for
1.2
laminar flow airfoils. However, low-speed models are often
equipped with trip strips that fix the transition point on the 0.8
model and may reduce the requirement for extremely low
0.4
turbulence. The need for low test-section turbulence is not
as severe for small student tunnels as it would be for
0.0
1.0
1.4
1.8
2.2
2.6
3.0
larger tunnels that are used for research and development
Turbulence Factor, TF
tests. The screens and honeycomb elements are effective Fig. 7 Variation of turbulence factor with
for reducing turbulence in wind runnels.
turbulence intensity from hot-wire
probe measurements
AE 2751
July-Nov. 2016
10

Calibration of Supersonic Tunnels: Mach Number
AE 2751
July-Nov. 2016

AE 2751
July-Nov. 2016
11

Whereas Mach numbers in nearsonic and transonic tunnels are usually determined
only on the tunnel centerline and at the tunnel walls, Mach numbers in supersonic
tunnels are usually obtained off the tunnel center lines as well. The reason is that much
larger non
non-uniformities
uniformities of flow are possible in supersonic nozzles because they can be
caused by shock waves. An average Mach number in the vicinity of a model is desired
for testing and data reduction purposes and the cross-sectional area survey will give a
better average Mach number than the centerline survey. If the centerline Mach number
distribution is constant or varies gradually but continuously, the centerline survey is
usually adequate. However, the characteristics of this distribution are not known when
the calibration is started.
AE 2751
July-Nov. 2016

Calibration of Supersonic Tunnels: Total Pressure
Pitot pressures are measured by using a simple device called a Pitot probe (tube).
The Pitot probe is simply a tube with a blunt end facing into the airstream. The tube
normally has an inside to outside diameter ratio of 0.5 to 0.8 (larger ratio is preferred
with smaller outside diameter) and a length aligned with the airstream of 15 to 20
diameters. The pressure orifice is formed by the inside diameter of the tube at the blunt
end. A rake of nine Pitot probes used for calibration traverses of a test section is shown
in Fig. 8. A Pitot probe is both simple to construct and accurate to use. It should always
have a squared-off entry and the largest practical ratio of hole diameter to outside
diameter. It is insensitive to angle of attack upto 10 deg. for an orifice diameter only
10% of the outside diameter and 15 deg. for an orifice diameter only 98% of the
outside diameter.
Calibration errors due to angle of attack and hole size within the above ranges are
much less than actual flow deviations found in any reasonable tunnel.
AE 2751
July-Nov. 2016
12

Fig. 8 A traversing rake of nine Pitot tubes
AE 2751
July-Nov. 2016

At this point we may make the clarification that an open-ended tube facing into the
airstream always measures the stagnation pressure (a term identical in meaning to
total pressure) is sees. Above M=1, the shock wave that forms ahead of the tube
means that it sees not the freestream stagnation pressure but the stagnation pressure
behind a normal shock. This new value is called Pitot pressure and in modern
terminology implies a supersonic stream, although there is no error in calling the
pressure so measured in a subsonic stream Pitot pressure. Pressures measured by
Pitot probes are influenced by very low Reynolds numbers based on the probe
diameter. This effect is seldom a problem in supersonic tunnels as a reasonably sized
probe will usually have will have a Reynolds number above 500 or 1000, which is the
range
g above which the trouble starts.
AE 2751
July-Nov. 2016
13

Calibration of Supersonic Tunnels: Static Pressure
Static pressures in a supersonic flow are much more difficult to measure than static
pressures in a subsonic flow and Pitot pressures in a supersonic flow. The static pipe
used for calibrating nearsonic and transonic tunnels is rarely used as its presence in the
tunnel affects the flow in the test section.
section It changes the area ratio of the nozzle by
subtracting from the effective throat and test section area, and it also interferes with the
expansion pattern required for the development of uniform flow.
While static pressure probes are not used extensively for calibrating supersonic tunnels, a
great deal of effort has been devoted to the development of accurate static pressures for
other applications. The result has been the development of probes in wind tunnel
calibration and use.
The primary problem in the use of static pressure probes at supersonic speeds is that any
probe will have a shock wave at its forward end which a rise in static pressure. If the probe
consists of a cone tip followed by a cylinder, the air passing the shoulder will be expanded
to a pressure below stream static pressure. Then as distance from the shoulder increased,
the pressure on the probe will approach the stream true static pressure.
AE 2751
July-Nov. 2016

Pressure measurements on a cone-cylinder probe with a 7-deg. included angle cone
followed by a cylinder 30 diameters in length are presented in Fig. 9. These results show
negligible errors in static pressure measurements for orifices located 10 diameters
downstream of the shoulder
shoulder.
Another type of probe designed for the accurate measurement od static pressures over a
large Mach number range is shown in Fig. 10. Errors in static pressures measured with
this probe are presented in Fig. 11. They were obtained by reference to a static pressure
calculated from a measured Pitot and total pressure. Flow angularity during these tests
was of no consequence, since the tunnel employed had extremely even flow. However it
was found impossible to get consistent results until the Pitot probe mentioned above and
the static probe were mounted in the test section on a vertically moving support system so
that either could be moved into centerline calibration position without a tunnel shutdown
and with little time interval between measurements. Evidently the extreme accuracy being
sought (of the order of 0.01% of q or 0.1% of p), minute tunnel changes due to controls or
thermal expansion became significant.
AE 2751
July-Nov. 2016
14

Fig. 9 Effect of orifice distance from shoulder on measured static pressure
AE 2751
July-Nov. 2016

Static pressures on the walls of supersonic tunnels are often used rough estimation of
test section Mach numbers. It is noted, however, that wall pressures do not necessarily
correspond to pressures on the centerline because of the possibility of compression or
expansion waves between the wall and the centerline. When Mach number is to be
determined from static pressure measurements, the total pressure of the stream is
measured in the settling chamber simultaneously with the test section static pressure.
Mach number is then calculated from the two pressures.
Fig. 10 Approximate dimensions of supersonic static pressure probe

There are three interdigitated rings of holes at a, b and c
AE 2751
July-Nov. 2016
15

Calibration of Supersonic Tunnels
Fig. 11 Error in static pressure measured pm as a fraction of the true static pressure
AE 2751
July-Nov. 2016

Results of calibrations to determine Mach number from tunnel wall static pressures and
tunnel axis Pitot pressures are presented in Fig. 12. This figure illustrates the difference
that may exist between the tunnel wall and centerline due to expansion waves between
the two.
two Results of a very thorough Mach number survey at one axial station of a tunnel
are presented in Fig. 13 and an even more complete survey along the plane of the
horizontal axis is shown in Fig. 14.
The importance of calibrating over the range of Reynolds numbers (pressures if the
temperature is constant) at which the tunnel will be operated is shown in Fig. 15. A change
in Reynolds number from 0.06 to 0.4x106 per inch causes a Mach number deviation of
0.06 at a nominal Mach number of 5. As mentioned previously, this effect results from
changes in boundary layer thickness and consequently effective nozzle area ratio.
As a matter of interest, many supersonic tunnels run at constant dynamic pressure
throughout their Mach number range. This is in contradiction with the test parameter, V2.
Constant q helps with handling balance loads, and with data reduction.
AE 2751
July-Nov. 2016
16

Fig. 12 Wall and tunnel axis calibration data from M=2 nozzle
AE 2751
July-Nov. 2016

Fig. 13 Contour plot of M=3 nozzle Fig. 14 Mach number distribution the AEDC M8 B tunnel
AE 2751
July-Nov. 2016
17

Mach
h number deviatiion, M
0.00
-0.08
Calibration of Supersonic Tunnels

M=1.5
M=2.5
M=3.5
M=4.5
M=5.0
M=5.5
-0.16
M=6.0
-0.24
0.1
0.2
-6
0.3
0.4
Reynolds number per inchX10

Fig. 15 Variation of centerline Mach number with Reynolds number
AE 2751
July-Nov. 2016

Calibration of Supersonic Tunnels: Flow Angularity
The flow angularity in a supersonic tunnel is usually determined by using either cone or
wedge yawmeters. The sensitivities of several yawmeters for use in supersonic tunnels
are presented in Fig. 16. The sensitivities of these yawmeters are maximum when the
wedge
d or cone angles
l maximum.
i
Th
They work
kb
below
l
M
Mach
h numbers
b
ffor which
hi h wave
detachment occurs, and are so used. The cone yawmeter is used more exclusively than
the wedge yawmeter because it is easier to fabricate and more robust. A photograph of
a wedge yawmeter is shown in Fig. 17.
The use of the yawmeter at supersonic speeds requires a calibration to determine the
aerodynamic error. A typical summary of results from tests to determine flow angularity
in a wind tunnel is presented in Fig. 18.
AE 2751
July-Nov. 2016
18

Fig. 16 Sensitivity of several yaw meters at supersonic speeds, pressure ratio per degree
AE 2751
July-Nov. 2016

Fig. 17 A wedge yawmeter
AE 2751
Fig. 18 Maximum up and down flow in a supersonic

tunnel for about 0.7 tunnel height
July-Nov. 2016
19

Calibration of Supersonic Tunnels: Turbulence Level
Measurements with a hot-wire anemometer Table 1: Turbulence in settling chamber and
demonstrate that there are high-frequency
test section of a supersonic tunnel
fluctuations in the airstream of supersonic
Settling Chamber Test Section
tunnels that do not occur in free stream.
4.5
Mach number
All
2.2
These fluctuations, broadly grouped under
1%
Sound, Pt/Pt
Less than 0.1%
0.2%
the heading, turbulence, consists of small
Entropy, Tt/Tt
Less than 0.1%
Less than 0.1%
oscillations in velocity, stream temperature
Vorticity,
V/V
0.5 to 1%
Less than 0.1%
(entropy) and static pressure (sound).
Values obtained from one tunnel are presented in Table 1.
The fluctuations arise from a variety of causes, mostly from the pressure regulator
valve, the drive system, the aftercooler, and the test section boundary layer. Velocity
fluctuations emanating from upstream causes may be reduced at low and moderate
Mach numbers by the addition of screens in the settling chamber. At high Mach
numbers, upstream pressure and velocity effects are usually less, since the large
nozzle contraction ratios damp them out. Temperature fluctuations are unaffected by
the contraction ratio. The existence of such fluctuations is of less interest than their
effect. Here the calibration procedure has been to determine the transition Reynolds
number on smooth cones and compare this value with values obtained in other tunnels.
AE 2751
July-Nov. 2016

Calibration of Supersonic Tunnels: Determination of Transition
By common usage, transition cones have either 5- or 10-deg included angles and highly
polished surfaces. Various methods have been used to determine the point of transition
on the cone. These have been (1) optical methods in which Schlieren pictures or
shadowgraphs
h d
h were made
d and
d iinspected
t d tto d
determine
t
i th
the point
i t att which
hi h a sudden
dd
thickening of the boundary layer occurred; (2) traversing along the cone a constant
AE 2751
July-Nov. 2016
20

By common usage, transition cones have either 5- or 10-deg included angles and
highly polished surfaces. Various methods have been used to determine the point
of transition on the cone. These have been (1) optical methods in which Schlieren
pictures or shadowgraphs were made and inspected to determine the point at
which a sudden thickening of the boundary layer occurred; (2) traversing along the
cone a constant distance away from the surface with a small Pitot probe that is
within the turbulent boundary layer and noting the point at which the probe
pressure changes from a steady to a fluctuating pressure; (3) making similar
traverses with a hot-wire anemometer and (4) measuring temperatures of the
surface by using thermocouples on the inner surface of a thin walled cone.
The dimensions of a cone for use with the surface temperature technique are
presented in Fig. 19. Surface temperatures of this cone were used to obtain a
recovery factor, Rr=(Taw-T)/(Tt-T).
where Taw=adiabatic wall temperature, OK Tt=stream total temperature, OK
and
T =stream static temperature, OK
The resulting data, which indicates the method of locating the point of transition is
shown in Fig. 20. A summary of transition Reynolds numbers in several wind
tunnels is presented in Fig. 21.
AE 2751
July-Nov. 2016

Fig 19 Dimensions of JPL transition cone

Fig.
AE 2751
July-Nov. 2016
21

Fig. 20 Typical determination of transition Reynolds number on a cone

Freestream Reynolds number per foot, 4.31x105, Freestream Reynolds number per foot, 4.31x105
AE 2751
July-Nov. 2016

Fig. 21 Transition Reynolds number on 5- and 10-degree cones as measured in several facilities
AE 2751
July-Nov. 2016
22

Calibration of Supersonic Tunnels: Test Section Noise
Test section noise is defined as pressure fluctuations. Noise may result from unsteady
settling chamber pressure fluctuations due to upstream flow conditions. It may also be
due to weak unsteady shocks originating in a turbulent boundary layer on the tunnel
wall Such weak shocks to which noise is attributed are illustrated by the shadowgraph
wall.
of Fig. 22. Noise in the test section of a wind tunnel is very likely to influence the point of
boundary layer transition on a model. Although it is possible that other effects on tests
results may exist, there has not been evidence showing that they do.
Test section noise can be detected by either hot-wire anemometer measurements or by
high-response Pitot pressure measurements. The method used to determine if the
noise is coming from the test section boundary layer is to make measurements in the
tunnel settling chamber as well as in the test section. It is then possible to determine if
fluctuations in the two places are related. It has been found that test section noise
generally increases as tunnel operating pressure increases, and, that test section noise
originating in the settling chamber generally decreases as tunnel Mach number increases.
AE 2751
July-Nov. 2016

Calibration of Supersonic Tunnels: Test Section Noise
Fig. 22 Noise emanating from the turbulent boundary layer on a missile model
M=3.5; Re=2x106/inch. Note the diminution of wavelet strength as the distance from the source is increased.
AE 2751
July-Nov. 2016
23

Calibration of Supersonic Tunnels: The Use of Calibration Results
AE 2751
July-Nov. 2016
24
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t U Tube
T b
The basic notion off pressure that is implemented in measurement
technology
gy is force p
per unit area. Other concepts
p such as those
embodied in various gas laws and theory are important in wind
tunnel work but are necessary in relating pressure to other ideas
and results rather than directly in the measurement of pressure
pressure.
One of the oldest devices for measuring pressures, and one of the
easiest to build,
build manometer; a term normally applied to a device
used to measure differential pressure.
pressure
Common applications
pp
are measurements of the difference
between a reference pressure such as atmospheric and a process
pressure such as a port on a wing and measurements of the
Fi 1 U-tube
Fig.
U t b manometer
t
difference between two pressures in a process
process, such pressures
f
from
th
the ttotal
t l and
d static
t ti ports
t off a Pit
Pitot-static
t t ti tube.
t b A simple
i l
manometer ((shown in Fig.
g 1)) can be made from
f
two pieces off
straight
g tubing
g made p
parallel and connected byy tubing
g at the
bottom or by bending tubing into a U shape.
shape The tubing is filled
with a liquid
liquid, and the difference heights in the two tubes is
measured usually by an attached scale.
measured,
scale
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t U Tube
T b
The tubing
g is filled with a liquid,
q , and the difference heights
g
in the two tubes is measured,,
usually by an attached scale.
scale The plane defined by the centerlines of the two tubes may
be inclined at varying angles to the horizontal
horizontal, which provides varying sensitivity.
sensitivity Setting
th ttubes
the
b vertical
ti l is
i the
th mostt common case. Th
The diff
difference in
i pressures is
i related
l t d to
t the
th
height difference and the parameters describing the manometer by the hydrostatic
pressure equilibrium relation:
p2-p
p1=h
h sin g (l-g)
where h is the difference in extents of liquid

q
columns in the two tubes and is the angle
g
between the horizontal and the plane of the parallel manometer tubes.
tubes
Th product
The
d t h
h sin
i is
i the
th vertical
ti l difference
diff
in
i the
th heights
h i ht off the
th two
t
columns.
l
Using smaller values of provides an increased sensitivity.
sensitivity Here g,
g is the weight per
unitit mass, also
l referred
f
d to
t as acceleration
l ti off gravity.
it The
Th value
l needed
d d here
h
is
i the
th
"local" value,, which varies with location on Earth and with altitude.
Also l is the density of liquid in the manometer (equivalently specific gravity times
Also,
densityy of water)) and g, is the densityy of air in most wind tunnel applications.
pp
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t U Tube
T b
The manometer
Th
t iis the
th mostt fundamental
f d
t l instrument
i t
t readily
dil available
il bl for
f pressure
measurement. Manometers are used frequently
q
y for calibrating
g and checking
g other
devices,, as it is difficult to obtain a more accurate or precise
p
result in the range
g of
differential pressures commonly of interest in subsonic aerodynamic testing.
In the past
past, wind runnels have used a wide array of manometers.
manometers These have from the
simple
i l U-tube,
U t b similar
i il to
t Fig.
Fi 4.1,
4 1 using
i a ruler
l or strip
t i off graph
h to
t measure flfluid
id heights
h i ht
to large
g banks of 50-200 tube manometers with constant level reservoirs to maintain the
reference fluid height
g to manometers that can be precisely
p
y read using
g some of the
devices. The U
U-tube
tube and precision manometers could typically provide one or two
pressures and the large multitube manometers were used to determine pressure
pressures,
distributions over model surfaces.
surfaces The simple,
simple precision
precision, and small multitube
manometers usually were read and recorded by an individual
individual. The large multitube
manometers
t
were photographed,
h t
h d and
d th
the fil
films were read
d later
l t by
b various
i
techniques.
t h i
The test
test-section
section total and static pressures were put on two of the manometer tubes.
Then the pressure coefficients were merely the h of the desired pressure divided by
the h of the dynamic pressure.
pressure To process data from large test programs,
programs the
photographs of the manometers were used with optical schemes arranged to scale the
images so that the pressure coefficients could be ready directly using a microfilm
machine.
hi
Thi could
This
ld be
b considered
id d a form
f
off optical
ti l data
d t processing.
i
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t M
Manometer
t
Two useful manometers that can be built in almost anyy laboratoryy are described. The
first is a simple unit with a few tubes and adjustable slant angle (Fig.
(Fig 2).
2) The second is a
unit (Fig.
(Fig 4.3)
4 3) capable of very precise measurements with h readings of 0.0001
0 0001 in.
in
b i possible
being
ibl if an appropriate
i t micrometer
i
t or height
h i ht gage is
i used
d in
i the
th construction.
t ti
To achieve repeatability of such fine measurement, it is necessary that the thermal
environment be quite stable.
stable For example,
example if the room temperature is around 75 OF,
F a
th
thermal
l shield
hi ld mustt b
be provided
id d between
b t
the
th person observing
b
i the
th meniscus
i
through
th
h
the optics
p
and the main part
p of the unit. Otherwise steady
y drift will be observed due to
the radiant heat from the person to the manometer unit and the corresponding
expansion of the material
material.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Multitube
M ltit b Inclination
I li ti M
Manometer
t
Fig 2 A multi tube variable inclination manometer

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Multitube
M ltit b Inclination
I li ti M
Manometer
t
Fig 3 A hi
Fig.
high
gh p
precision
i i micro
i
manometer
t
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Pressure
P
T
Transducers
d
The most commonly used pressure transducers are of the diaphragm type,
type which simply
means that the basic sensing mechanism is a thin sheet of material that deforms as the
pressure across it changes.
changes The methods of sensing the diaphragm deformation include
strain gages directly attached to the diaphragm,
diaphragm circuits to sense the change in capacitance
due to the geometric change
change, and circuits to sense the change in inductance due to the
geometric change.
change The strain gage units are most numerous as they can be made smaller
and
d with
ith currentt solid-state
lid t t electronic
l t i methods
th d can be
b made
d ffor a ffew d
dollars.
ll
Th
These units
it
are remarkable
k bl for
f their
th i economy although
lth
h they
th mustt b
be calibrated
lib t d frequently
f
tl against
i t more
stable standard units.
There are many strain gage units with stainless steel diaphragms and even some quartz
diaphragms. Transducers using the capacitive sensing method tend to be the higher priced
ranges and to have larger dynamic range capability as well as being more rugged than the
strain-gage-based units.
units These are more likely to be found the main tunnel condition
sensing applications and for secondary calibration
calibration.
A majority of the pressure transducers used in subsonic wind tunnels are of differential type
type,
with
ith a selected
l t d reference
f
pressure applied
li d to
t the
th reference
f
side.
id Th
The absolute
b l t pressure
versions
i
are almost
l
t always
l
available.
il bl They
Th differ
diff from
f
the
th differential
diff
ti l units
it by
b having
h i one
side
id off the
h di
diaphragm
h
permanently
l sealed
l d to a fixed
fi d pressure that
h iis close
l
to vacuum. This
Thi
is necessaryy in order to avoid serious temperature
p
sensitivityy and leads to the absolute units
having
g diaphragms
p g
that can withstand least 1 atm. It turns out that this means theyy are not
sufficiently
su
c e y se
sensitive
s e for
o app
applications
ca o s in low-speed
o speed wind
d tunnels.
u es
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Micro
Mi
manometer
t
The output from the pressure transducers can be read by analogue or digital meters.
meters A
digital high precision micro manometer is shown in Fig.
Fig 4.
4
Fig 4 A digital high precision micro manometer (FCO560)

Fig.
AE 2751
July-Nov.
July Nov 2016
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Scanning
S
i g Box
B
For measurement of multiple
p
pressures,
p
essu es, from
o tthe
e stat
static
cp
pressure
essu e
taps on the models,
models a scanning
(selection) box (Fig.
(Fig 5) or a
scanivalve along with a digital high
precision
i i micro
i
manometer
t can be
b
used. Each scanning
g box can be
connected to 20 nos. of static
pressure taps.
taps Slave units with 20
nos can be added to increase the
nos.
number
b off static
t ti pressures to
t be
b
measured.
Fig 5 A 20-channel scanning box

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Pressure
P
Scanner
S
An alternative to the
scanning box and digital
hi h precision
high
i i micro
i
manometer is a pressure
p
scanner (Fig. 6). A
pressure scanner has 32
individual pressure
transducers The range of
transducers.
each
h off th
these ttransducers
d
can be different. Additional
scanners each of 32
channels can be added to
increase the number of
pressures to
t be
b measured.
d
The size of the scanner is
sufficiently to be mounted
inside the wind tunnel
model.
model
AE 2751
Fig 6 A pressure scanner (Scanivalve ZOC23B/32Px)

Fig.
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Pressure
P
Sensitive
S
iti P
Paint
i t
Almost all of the preceding pressure
pressure-measuring
measuring devices require that the pressure at a point
off iinterest
t
t be
b communicated
i t d to
t the
th sensing
i element
l
t from
f
II pressure "tap"
"t " or "port"
" t" through
th
h
a small tube to the sensor. There are two major
j p
problems with this method. The most
important one is that it makes the models quite expensive to build.
build The second is that the
time response of the pressure measurement is limited by the presence of the tubing as a
transfer mechanism. There have been a number of important
p
experiments
p
using
g surfacemounted transducers, but these are by far the most expensive models and test programs.
In any case
case, there can never be a large enough number of pressure taps to provide high
accuracy for
f forces
f
obtained
bt i d by
b integration
i t
ti off measured
d surface
f
pressure.
Pressure-sensitive paint offers the promise of very high spatial density of measurements
with a moderate impact on model cost.
cost There is,
is at present,
present a high capital cost for the
cameras needed for high-quality
g q
y results. This new method is being
g used extensivelyy in
transonic testing but has only recently been demonstrated to be capable of useful results
for test speeds as low as 100 knots.
knots The PSP principle is schematically illustrated in Fig.
Fig 7.
7
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
P
Pressure
M
Measurements:
t Pressure
P
Sensitive
S
iti P
Paint
i t
Fig. 7 Oxygen
Fig
Oxygen-quenched
quenched photoluminescence process
AE 2751
July Nov 2016
July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t P
Probes
b
A Pitot tube is used to measure total pressure.
p
The shape
p of the tube affects its
sensitivity to flows inclined to the tube axis.
axis Pitots with hemispherical noses begin to
show errors in total pressure at very low angles of flow inclination.
inclination Pitot tubes with a
sharp
h
square nose begin
b i to
t show
h
errors near 8O flow
fl
inclination.
i li ti
This
Thi can be
b improved
i
d
by chamfering the nose.
A Kiel
Ki l tube
t b can provide
id accurate
t stagnation
t
ti pressure for
f flow
fl
angles
l beyond
b
d 30O. Other
Oth
tip
p shapes
p p
provide various sensitivities,, as indicated in Fig.
g 8. It is a relativelyy easyy task
to measure a Pitot's
Pitot s sensitivity to flow angle by use of a flow of known angularity.
angularity The
Pitot is pitched or yawed depending on which is more convenient to determine its
sensitivityy to fflow angularity.
g
y
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t P
Probes
b
Fig. 8 Variation of measured stagnation pressure with yaw for selected probe
geometries: Cpa=2[p(a)-p(0)]/V2
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t P
Probes
b
As previously indicated,
indicated a total-head
total head tube with a hemispherical tip will read the total
head accuratelyy as long
g as the yyaw is less than 3O. A squared-off
q
Pitot tube will go
g to
higher angles without error,
error but both square
square- and round-tip
round tip Pitot tubes suffer errors if
th are used
they
d att ttoo llow R
Reynolds
ld numbers
b
or ttoo close
l
to
t a wall.
ll
At very low Reynolds numbers,
numbers the flow regime is referred to as "creeping
creeping flow
flow." The
difference
ff
between the pressure and stagnation
g
pressure is not the dynamic
y
pressure,
2
q=00.55V , which gives cpstag
1 for high
high-Reynolds-number
Reynolds number flow.
flow In this regime the
t =1
stagnation press
pressure
re on a bl
blunt-nosed
nt nosed bod
body is given
gi en approximately
appro imatel in terms of the
Reynolds
y
number based on the bodyy diameter as
cpstag=((pstag-pp)/q=1+6/Re
1 6/R D
Corrections for Pitot tubes under these conditions are shown in Figs.
Figs 9 and 10.
10
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t P
Probes
b
1.3
Cppsttaggnn
1.2
Square tip
Hemisphere tip
1.1
1.0
10
100
1000
Reynolds number based on probe tip diameter,
diameter ReD
Fig 9 Performance of Pitot probes at low Reynolds numbers

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t P
Probes
b
Fig 10 Velocity correction for a circular Pitot probe near a wall

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
The most common device for determining the total pressure or total head and the static
pressure of a stream is the p
p
pilot-static lube, an instrument that yyields both the total head
and the static pressure
pressure. A "standard"
standard Pitot-static
Pitot static tube is shown in Fig.
Fig 11.
11 The orifice at
A senses total
t t l head,
h d p+0.5
+0 5V2, and
d th
the orifices
ifi
att B sense the
th static
t ti pressure. This
Thi
discussion is limited to subsonic flows. If the pressures
p
from the two orifices are
connected across a manometer or pressure transducer
transducer. the pressure differential will be
approximately
pp
y 0.5V2, from which the velocityy mayy be calculated provided
p
the densityy
is available.
available The density can be calculated from the equation of state based on a
temperat re meas
temperature
measurement
rement and the static press
pressure
re meas
measurement.
rement
The Pitot
Pitot-static
static tube is easy to construct,
construct but it has some inherent errors.
errors If due
allowance
ll
i made
is
d for
f these
h
errors, a determination
d
i i off the
h dynamic
d
i pressure within
i hi
about 0.1%
0 1% is possible.
possible
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
Fig 11 Pitot-static
Fig.
Pitot static probe
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
The pressure sensed at the "static"
static holes differs from the stream static pressure due to
t
two
effects
ff t off th
the geometry.
t Th
The fifirstt effect
ff t iis generic
i for
f a semi-infinite
i i fi it axisymmetric
i
ti
bodyy with flow approaching
pp
g the "nose" of the bodyy along
g the axis of symmetry.
y
y The nose
of the probe has a region on the upstream surface where the flow stagnates and the
pressure is above the stream pressure of the approaching flow.
flow At the stagnation point,
point
the p
pressure is the total pressure
p
of the stream. The flow accelerates from the
stagnation point around the curved surface and the local surface pressure rapidly falls
through and below the stream or static pressure in the approaching flow.
flow A minimum
pressure is
i reached
h d somewhere
h
on the
th curved
d surface,
f
and
d the
th pressure starts
t t to
t
"recover"
eco e to
toward
a d the
t e stream
st ea stat
staticc pressure
p essu e as one
o e moves
o es downstream
do st ea along
a o g the
t e
surface parallel to the stream direction.
direction If the probe were infinitely long and aligned with
th flflow, th
the
then the
th pressure on the
th ttube
b surface
f
would
ld asymptotically
t ti ll approach
h the
th
stream p
pressure with distance from the nose.
This effect means that pressure taps on the probe's
probe s parallel surface a finite distance
f
from
th
the nose would
ld produce
d
a measurementt th
thatt iis llower than
th the
th llocall stream
t
pressure. The amount of the difference in indicated in Fig. 12.
AE 2751
July Nov 2016

July-Nov.
10
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
Fig 12 Effect of static orifice distance from tip or stem

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
The second effect is associated with the presence of the "stem
stem," a cylinder whose axis
is perpendicular to the approaching stream.
stream A high-pressure
high pressure region exists ahead of the
stem and that region includes the surface of the probe itself.
stem,
itself In fact,
fact the Prandtl design
h a stagnation
has
t
ti point
i t near the
th intersection
i t
ti off the
th stem
t
and
d the
th probe
b body.
b d This
Thi effect
ff t
creates
t pressures on partt off the
th probe
b surface
f
that
th t are higher
hi h than
th the
th static
t ti pressure
in the approaching
pp
g stream.
The two effects may cancel each other if the static hole locations are properly chosen.
chosen
The "standard"
standard Pitot
Pitot-static
static tube does not employ this principle because the static hole
l
location
ti is
i so critical
iti l that
th t smallll deviations
d i ti
in
i construction
t ti or damage
d
to
t the
th tip
ti can
produce
d
a relatively
l ti l large
l
error in
i the
th static
t ti reading.
di
Th
The P
Prandtl
dtl d
design
i (Fig.
(Fi 13) is
i
intended to take advantage
g of this cancellation.
If a new Pitot-static tube is to be built,
built either it may be designed as per Fig.
Fig 11 and its
static pressure readings corrected as per Fig.
Fig 12 or the Prandtl design may be used
used.
The Prandtl design (Fig.
(Fig 13) should require no correction but should be checked for
accuracy. E
Existing
i ti Pit
Pitot-static
t t ti tubes
t b should
h ld be
b examined
i d for
f tip
ti and
d stem
t
errors so that
th t
their constants may
y be found.
If a long static tube is available
available, the static pressure can be determined along a
longitudinal line in the test section.
section Then the Pitot
Pitot-static
static tube can be placed on this line
and the static pressure orifices in a Pitot-static
Pitot static tube can be calibrated
calibrated.
AE 2751
July Nov 2016

July-Nov.
11
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
If a long static tube is available,
available the static pressure can be determined along a
l
longitudinal
it di l line
li in
i the
th test
t t section.
ti
Then
Th the
th Pitot-static
Pit t t ti ttube
b can b
be placed
l
d on this
thi line
li
and the static p
pressure orifices in a Pitot-static tube can be calibrated.
The accuracy of a standard Pitot-static
Pitot static tube when inclined to an airstream is shown in
Fig.
g 14. If a Pitot-static tube is p
placed near a model, the model's static p
pressure field will
influence the pressure sensed by the static ports
ports, and the reading will not be the free
freestream velocity
velocity. This is why tunnel dynamic pressure calibrations are made without a
model
d l in
i the
th test
t t section.
ti
The
Th same problem
bl
exists
i t on an aircraft,
i
ft where
h
greatt care has
h
to be ta
taken
e in the
t e location
ocat o of
o a static
stat c pressure
p essu e sou
source
ce so as to find
d a location
ocat o where
e e the
t e
static pressure varies as little as possible with lift or aircraft attitude.
attitude
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
Fl
Flow
Instrumentation:
I t
t ti
Pitot
Pit t Static
St ti Probes
P b
Fig 13 Prandtl
Fig.
Prandtls
s design for a Pitot
Pitot-static
static probe
AE 2751
Fig 14 Performance of the standard PitotFig.

Pitot
static probe in yaw
July Nov 2016

July-Nov.
12
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements
t
The flflow velocity
Th
l it can be
b measured
d using
i one off the
th following
f ll i techniques:
t h i
1 Pressure probes
1.
2. Pressure probes
p
using
g fast response
p
pressure
p
transducers
3 Thermal Anemometry
3.
Anemometr
4 Optical Anemometry: Laser Velocimetry
4.
Velocimetry, Laser Doppler Anemometry and Particle
Image Vecocimetry
The above techniques can measure the three velocity components in an incompressible
flow For measurements in compressible flow,
flow.
flow additionally temperature has to be
measured
d using
i a suitable
it bl instrument.
i t
t
Pressure probe measures time averaged velocity and its components.
components Pressure probes
using
sing fast response press
pressure
re transducers,
transd cers thermal anemometr
anemometry and optical
anemometry
y measures time dependent
p
velocityy and its three components.
p
In addition
the six Reynolds stresses can also be measured.
measured
F th
For
the sake
k off b
brevity,
it lilimited
it d discussion
di
i off the
th above
b
techniques
t h i
is
i presented.
t d
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Pressure
P
Probes
P b
Yawheads: A yawhead is a sphere that has two or more static ports on the forward face
of the sphere (Fig.
(Fig 12).
12) If the ports are at exactly 45O to the centerline of the support
and the flow is parallel to the support,
support then (Pa-P
Pb)/q=P/q=0.
=0
If there
th
iis flflow angularity,
l it then
th P/q0,
0 and
d the
th value
l off P/q will
ill be
b a function
f
ti off the
th
flow angle.
g In practice,
p
a yyawhead must be calibrated. The probe
p
is placed
p
in a flow that
has no angularity,
g
y, and the probe
p
is pitched
p
or yawed
y
about its center through
g an angle
g
range
g both p
positive and negative.
g
This is done for the y
yawhead in the "upright,"
p g , or
normally used, position, and then the yawhead is rotated 180O to an "inverted"
inverted position.
This wilI result in two curves of P/q=0
0 (upright and inverted) versus the angle for the
yawhead If the static ports are symmetrical to the support axis
yawhead.
axis, the two curves will lie
on top of each other
other. If there is an asymmetry in the static port locations or possibly in
the conditions of the static ports themselves,
themselves the curve will be displaced by twice the
error The true curve lies halfway between the two measured curves.
error.
curves If the flow used to
calibrate the yawhead is not parallel to the yawhead support axis at the zero angle,
angle the
tr e ccurve
true
r e will
ill not pass through
thro gh the zero
ero angle
angle. It is normally
normall desired to have
ha e the yaw
a
h d calibration
head
lib ti independent
i d
d t off dynamic
d
i pressure and
d static
t ti pressure so the
th coefficient
ffi i t
d fi d b
defined
below
l
may be
b used:
d
Cy=(p
(pb-p
pa))/[p
[pc-0.5(p
(pa+p
pb)]
It is instructive to investigate the variation of the coefficient as if the pressure distribution
on the spherical yaw head were accurately given by potential flow theory for a sphere in
a uniform stream
stream.
AE 2751
July Nov 2016

July-Nov.
13
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Pressure
P
Probes
P b
Fig
Fig.
g 15 Yawhead
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Pressure
P
Probes
P b
Quite
Q
ite often a yawhead
a head has five
fi e static ports rather than the three just
j st described.
described With
such an arrangement,
g
, the two flow angles
g
required
q
to define the three-dimensional flow
vector may be obtained although the calibration is now over the two angles and a
surface must be generated rather than a single curve
curve. Pitot
Pitot-static
static tubes that use a
h i h i l nose can be
hemispherical
b made
d to
t have
h
two
t
yawheads
h d built
b ilt into
i t the
th nose. In
I this
thi case,
one instrument at any place in the test section will measure total pressure, static
pressure dynamic pressure,
pressure,
pressure upflow,
upflow and cross-flow.
cross-flow This is quite useful for determining
fl
flow
conditions
diti
in
i a test
t t section.
ti
Claw p
probe: A claw probe
p
will also measure flow angularity
g
y and is simple
p to build. In its
simplest form a claw probe consists of two parallel pieces of tubing that are bent 45
45
away from their common axis and then bent back 90
90 toward their common axis (Fig.
(Fig
16) The
16).
Th two
t
heads
h d off the
th probe
b are cutt off
ff square about
b t two
t
diameters
di
t
from
f
the
th
centerline. Often a third tube is added to measure total pressure, and the two claws can
be made to simultaneously measure both cross-flow
cross flow and upflow
upflow. The calibration
t h i
technique
for
f a claw
l
probe
b is
i the
th same as that
th t for
f a yawhead.
h d Cl
Claw probes
b are more
delicate than yyawheads because the two tubes used to measure can be easily
y bent,,
thus changing the calibration
calibration.
AE 2751
July Nov 2016

July-Nov.
14
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Pressure
P
Probes
P b
Fig 16 Cl
Fig.
Claw p
probe
b
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Pressure
P
Probes
P b
Cone Probes:
C
P b
C
Cone
probes
b with
i h five
fi or seven boles
b l are widely
id l used
d for
f flow
fl
vector
measurements. They
y have the advantage
g of simplicity
p y of construction and can be calibrated
more readily at higher angles.
angles Gupta gives detailed treatment and examples of calibration
results Gallington and Gerner et al
results.
al. give results that allow calibration for compressible flow.
flow
Probes for Measurements in Reverse Flow: A number of developments
p
have been
carried
ca
ed out o
on probes
p obes to allow
a o measurements
easu e e ts to be made
ade in flow
o with
t angles
a g es from
o almost
a ost
any direction.
direction Cogotti describes and has used a 14
14-hole
hole probe extensively
extensively. This probe is
essentially two 7-hole
7 hole cone probes in a back-to-back
back to back configuration
configuration.
Yamaguchi
g
et al. and Rediniotis have developed
p spherical
p
probes
p
with 13 holes arranged
g as
multiple 5
5-hole
hole yaw probes. Gupta gave procedures for a spherical probe with 11 holes
arranged in the pattern of the faces of a dodecahedron with the probe stem emerging from
the location of the 12th face center.
center
Other Pressure Devices: Sometimes a simpler
p version of a y
yawhead is used. One device
consists of five tubes arranged in a cross configuration with one tube in the center and two
pairs of tubes attached to it at right angles.
angles The center tube is cut off with its end
approximately perpendicular to the flow.
flow The other tubes are cut off at a 45O angle.
angle A
second
dd
device
i is
i similar
i il but
b t consists
i t off two
t
parallel
ll l tubes
t b cutt off
ff att an angle.
l Both
B th off these
th
devices are calibrated in a manner similar to a y
yawhead or claw probe.
p
AE 2751
July Nov 2016

July-Nov.
15
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Devices
D i
R
Responding
p di g to
t Aerodynamic
A dy
i Forces
F
The devices described here are small
airfoils wedges
airfoils,
wedges, and vanes.
vanes They provide
a response based on aerodynamic forces
over surface
f
areas llarger th
than a typical
t i l
pressure port.
Airfoils
Ai
f il and
d Vanes:
V
Fl
Flow
angles
l can be
b
measured using
g small wings
g attached to
sting balances
balances. The flow angle calibration
Fi 17 Vane-type
Fig.
V
t
angularity
l it probe.
b U
Uses a fivefi
is obtained by the rotation of the drag
component
p
balance and can resolve about
polar.
l S
Smallll vanes off various
i
configurations
fi
i
0 006O. Owing to its natural high frequency
0.006
frequency,
can also be used (Fig. 17). Often these
probe can make continuous motion surveys
are attached to low-friction
low friction potentiometers
t read
to
d th
the angle
l or to
t balances.
b l
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Thermal
Th
l Anemometry
A
ty
Hot wires and hot films are used to obtain fast-response velocity measurements in
turbulent flows as well as mean velocities and,
and with multiprobes,
multiprobes flow angularity.
angularity The
probe or sensor is a fine wire (diameter
p
(
of a few micrometers)) or a cooled quartz
q
fiber
attached between two supporting needles on the probe (Fig.
(Fig 18).
18) Current passed through
the wire or film raises its temperature above the adiabatic recovery temperature of the
gas. The
Th hot
h t wire
i responds
d to
t changes
h
in
i total
t t l temperature
t
t
and
d mass flux
fl (To and
d U).
U) IIn
subsonic applications
pp
where the densityy is high
g and the fluid temperature
p
is low and
constant the problem of beat transfer through the support needle (end losses) and
constant,
radiation effects can be ignored,
ignored and the wire's
ire's response basically
basicall is a function
f nction of velocity
elocit
alone. Under these conditions and using
g appropriate
pp p
calibration and measurement of the
voltage across the bridge
bridge, both mean and turbulent velocities are obtained.
obtained
Electronic
El
t i hot-wire
h t i circuits
i it include
i l d a feedback
f db k system
t
to
t maintain
i t i the
th wire
i att a constant
t t
wire resistance. This is a constant
constant-temperature
temperature anemometer. Feedback can also be used
to maintain a constant-current anemometer
anemometer. Constant-temperature anemometers are
easiest
i t tto use iin subsonic
b
i or iincompressible
ibl flow,
fl
while
hil the
th constant
t t currentt anemometers
t
are p
preferred in compressible
p
flow. Frequency
q
y response
p
to 50 kHz is easilyy obtained.
AE 2751
July Nov 2016

July-Nov.
16
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Thermal
Th
l Anemometry
A
ty
Hot wires and hot films are used to obtain fast-response velocity measurements in
turbulent flows as well as mean velocities and,
and with multiprobes,
multiprobes flow angularity.
angularity The
probe or sensor is a fine wire (diameter
p
(
of a few micrometers)) or a cooled quartz
q
fiber
attached between two supporting needles on the probe (Fig.
(Fig 18).
18) Current passed through
the wire or film raises its temperature above the adiabatic recovery temperature of the
gas. The
Th hot
h t wire
i responds
d to
t changes
h
in
i total
t t l temperature
t
t
and
d mass flux
fl (To and
d U).
U) IIn
subsonic applications
pp
where the densityy is high
g and the fluid temperature
p
is low and
constant the problem of beat transfer through the support needle (end losses) and
constant,
radiation effects can be ignored,
ignored and the wire's
ire's response basically
basicall is a function
f nction of velocity
elocit
alone. Under these conditions and using
g appropriate
pp p
calibration and measurement of the
voltage across the bridge
bridge, both mean and turbulent velocities are obtained.
obtained
Electronic
El
t i hot-wire
h t i circuits
i it include
i l d a feedback
f db k system
t
to
t maintain
i t i the
th wire
i att a constant
t t
wire resistance. This is a constant
constant-temperature
temperature anemometer. Feedback can also be used
to maintain a constant-current anemometer
anemometer. Constant-temperature anemometers are
easiest
i t tto use iin subsonic
b
i or iincompressible
ibl flow,
fl
while
hil the
th constant
t t currentt anemometers
t
are p
preferred in compressible
p
flow. Frequency
q
y response
p
to 50 kHz is easilyy obtained.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Thermal
Th
l Anemometry
A
ty
Fig 18 Three
Fig.
Three-element
element hot-film
hot film probe
AE 2751
July Nov 2016

July-Nov.
17
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Optical
Opti l Anemometry
A
ty
A laser
l
velocimeter
l i t uses a tracer
t
method,
th d which
hi h depends
d
d on detection
d t ti off an optical
ti l
effect from a p
particle that is carried with the flow. The most common laser Doppler
pp
velocimeter (LDV) uses optics to split the laser beam into two parallel beams that
are focused to cross at the point where measurements are to be made (a dual-beam
dual beam
s stem) O
system).
Owing
ing to wave
a e interference
interference, a fringe pattern in an ellipsoid
ellipsoid-shaped
shaped volume
ol me
at the beam intersection is formed. A second lens assemblyy (the
(
receiver)) with a
small aperture is focused on this fringe region to collect light from seed particles
crossing the fringes
fringes. This light is fed to a photodetector that is used as the input to
the sophisticated electronic signal processor that measures modulated frequency
frequency.
In a dual-beam system
y
the frequency
q
y from the scattered light
g from the two
beams are superposed on the photodetector's
photodetector s surface
surface. The mixing process in the
photodetector then gives the difference in the frequency from the two beams
beams. All
other
th frequencies
f
i are too
t hi
high
h to
t detect
d t t (thi
(this iis called
ll d an optical
ti l heterodyne).
h t d
) If is
i
the angle
g between one of the beams and the bisector of the two beams and is the
wavelength of the laser light
light, then it can be shown that when the particle velocity is
much less than the speed of light the modulator frequency is
fD=2us sin
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Optical
Opti l Anemometry
A
ty
where us, is the velocity parallel to the plane of the two beams and perpendicular to the
bi
bisector
t off th
the b
beams. Th
Thus th
the relationship
l ti
hi between
b t
the
th flow
fl
velocity
l it and
d fD is
i linear
li
and a function of half the beam angle
g and the wavelength
g of the laser light.
g Perhaps
p it is
easier to think of the system working in the following manner:
When a p
particle moving
g with the flow passes
p
through
g each fringe,
g it is illuminated, which
causes a series of pulses from the photomultiplier.
photomultiplier As the distance between the fringes
is known,
known the time to cross the fringes is measured,
measured and this yields the particle velocity.
velocity
It should
h ld be
b noted
t d that
th t th
the ffringe
i
spacing
i iin micrometers
i
t
is
i equall to
t the
th calibration
lib ti
constant in meters per second per megahertz. When a large number of samples are
taken the signal processor and computer calculates both the average velocity and
taken,
i t t
instantaneous
velocity,
l it which
hi h can be
b used
d to
t obtain
bt i the
th turbulence
t b l
or velocity
l it variation.
i ti
AE 2751
July Nov 2016

July-Nov.
18
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Total
T t lH
Head
d Rake
R k
Airfoil
Ai
f il profile
fil drag
d
has
h often
ft been
b
measured
d by
b the
th use off a drag
d
wake
k rake.
k In
I this
thi
t h i
technique
the
th momentum
t
lloss iin the
th wake
k is
i determined
d t
i d by
b measurements
t with
ith a bank
b k
of total head tubes. The rake also should have two or more static tubes offset from the
total head tubes to obtain the local static pressure.
p
The rake size must be adequate
q
to
encompass
p
more than the width of the wake. Often,, the tube spacing
p
g is greater
g
at both
ends of the rake than in the center by a factor of 2. The spacing of the tubes must be
known with precision so that the momentum profile can be accurately determined.
determined
Generally the total tubes are made of 0.0625
0 0625-in
in. thin-wall
thin wall tubing (Fig
(Fig. 19).
19) The static
tubes must be offset from the plane of the total tubes to avoid interference effects on the
static pressure
pressure. Their purpose is to determine the static pressure in the wake.
wake Because
the static pressure can be affected by both the total head tubes and the base of the
rake they must be carefully calibrated so that the error in static pressure is known.
rake,
known
Since the static ttubes
bes m
must
st ha
have
e a hemispherical nose shape,
shape Krause
Kra se has shown
sho n that it
i possible
is
ibl to
t adjust
dj t this
thi shape
h
to
t reduce
d
the
th error to
t zero.
As an alternative to a wake rake, a mechanical traversing mechanism can be used.
Using various encoders, the location of a probe can be determined with a high degree
of accuracy
accuracy. The sensor can be a Pitot and,
and preferably,
preferably a static tube,
tube a multihole probe,
probe
a hot wire
wire, or a thin film or even a fiber-optic laser velocimeter head
head. When measuring
the momentum loss in the wake by any method,
method care must be taken to ensure that the
whole width of the wake is measured.
measured
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Total
T t lH
Head
d Rake
R k
Fig 19 T
Fig.
Total
t lh
head
d rake.
k N
Note
t single
i gl static
t ti p
pressure tube.
t b
AE 2751
July Nov 2016

July-Nov.
19
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t B
Boundary
d yL
Layer
y M
Mouse
Measurements in the boundaryy layer
y are often made to detect the transition between
laminar and turbulent flow or to find the local skin
skin-friction
friction coefficient
coefficient. Obtaining knowledge
of the velocity profile in the boundary layer is important in attempts to determine these
quantities.
titi
Withi
Within an attached
tt h d b
boundary
d
llayer the
th static
t ti pressure is
i essentially
ti ll constant
t t
while the total pressure varies. There are several ways in which the velocity profile can be
obtained.
obtained
The oldest method is byy use of a boundaryy layer
y mouse (Fig.
( g 20).
) This device is a series
of total head tubes
tubes, often with oval or flat inlets.
inlets To obtain the velocity profile with adequate
resolution at the surface requires the total head tubes to be spaced closer together than
their diameters. Thus, the total head tubes are placed on an inclined plane to obtain the
required close vertical spacing. The boundary layer mouse often has a static orifice to
measure the static pressure or the static pressure can be measured by a surface port.
port
D i use the
During
th mouse is
i attached
tt h d to
t the
th model.
d l The
Th boundary
b
d
layer
l
mouse measures the
th
velocityy profile
p
over a finite span
p of the model,, rather than a single
g spanwise
p
station.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t B
Boundary
d yL
Layer
y M
Mouse
Fig 20 Boundary layer mouse

Fig.
AE 2751
July Nov 2016

July-Nov.
20
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t T
Traversing
i g Probes
P b
The velocityy profile
p
can also be measured byy using
g a traversing
g mechanism whose
position off the surface can be quite accurately determined by a digital optical encoder.
encoder
The traverse mechanism can carry a single total head probe,
probe a hot wire or a split film.
film Very
good
d agreementt h
has been
b
shown
h
between
b t
Pitot
Pit t probes,
b
hot
h t wires,
i
and
d thin
thi films
fil
when
h
supported by a traverse mechanism.
IIn general,l it is
i better
b tt to
t supportt the
th boundary
b
d
layer
l
mouse or the
th traverse
t
mechanism
h i
from the model rather than the tunnel walls. This avoids two problems.
p
First,, when the
walls are used for support,
support the probes must be moved when the model is pitched and then
reset to obtain a very close and accurately known proximity to the surface.
surface The second
problem with a wall support is that most models tend to move slightly
g y and often
f
oscillate
when under loads owing to balance deflections. Or simply structural flexibility. If the
probes are being used to detect transition between laminar and turbulent flow,
flow extreme
care mustt b
be taken
t k to
t ensure that
th t th
the probe
b it
itselflf is
i nott causing
i transition
t
iti prematurely.
t l
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Transition
T
iti
Manyy methods are in use to determine the location of the transition region.
g
Theyy
include the following:
1 Pl
1.
Plott the
th velocity
l it gradient
di t in
i the
th boundary
b
d
layer
l
and
d determine
d t
i whether
h th the
th flow
fl
is
i
laminar or turbulent by the slope of the gradient, as illustrated in Figs. 21 and 22.
2 Determine
2.
D t
i the
th b
beginning
i i off transition
t
iti as the
th point
i t where
h
the
th velocity
l it as a function
f
ti off
streamwise distance at a fixed small height
g above the bodyy surface is a minimum,, as
illustrated in Fig.
Fig 23.
23
3. Read the static p
pressure at a small height
g above the surface, determine the transition
by a slight dip in the plot of static pressure versus percent chord.
chord
4 R
4.
Read
d the
th velocity
l it att a smallll height
h i ht above
b
the
th surface
f
with
ith a hot-wire
h t i anemometer
t and
d
note the transition as a region of unsteadiness in the output.
5 R
5.
Read
d the
th velocity
l it att a smallll height
h i ht above
b
the
th surface
f
with
ith a hot-wire
h t i anemometer
t or
thin-film gage
g g and note the start of transition as the point
p
of minimum velocity.
y
AE 2751
July Nov 2016

July-Nov.
21
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Transition
T
iti
Fig 21 Velocity distribution in laminar and turbulent boundary layers

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Transition
T
iti
Fig 22 Boundary layer six chordwise stations along an airfoil

Fig.
AE 2751
July Nov 2016

July-Nov.
22
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Transition
T
iti
Fig 23 Velocity in boundary layer at a constant small height above surface

Fig.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Transition
T
iti
6. Carefullyy emit smoke from flush orifices and note the transition byy the dispersal
p
of the smoke stream (may be difficult at high velocities)
velocities).
7. P
7
Paint
i t th
the model
d l with
ith special
i l chemicals
h i l that
th t evaporate
t slowly.
l l The
Th evaporation
ti
will proceed most rapidly where the flow is turbulent.
8. Li
8
Listen
t tto th
the b
boundary
d
llayer with
ith an ordinary
di
doctor's
d t ' stethoscope
t th
connected
t d
to a flat total head tube,, moving
g the total head tube p
progressively
g
y along
g the
surface from the beginning of the boundary layer in the downstream direction.
direction
As long
g as the flow
f
is laminar, a soft
f sh-sh-sh-sh can be heard. When it is turbulent, a
distinct roar is heard. This same input fed into a transducer becomes quite graphic on an
oscilloscope or amplified and fed to an audio speaker.
speaker
AE 2751
July Nov 2016

July-Nov.
23
Wind Tunnel
T nnel Measurements
Meas rements
V l ity M
Velocity
Measurements:
t Preston
P t Tubes
T b
This device is used to experimentally
p
y measure the wall coefficient of skin friction byy
measuring both a static pressure and total pressure at the same chordwise location.
location The
total pressure is measured by a Pitot tube that touches the surface
surface. This can be done
b
because
ffor unseparated
t d tturbulent
b l t flflow th
there iis a region
i near the
th wallll on the
th order
d off
10% of the boundary layer thickness in which the flow depends on the local wall skin
friction w, the density , the kinematic viscosity and a length.
length Preston took one-half of
th Pit
the
Pitot-tube
t t b di
diameter
t as the
th length.
l
th Dimensional
Di
i
l analysis
l i leads
l d to
t th
the equation:
ti
Other methods using thermal anemometry and optical anemometry are available to
measure skin friction and its temporal variation on the airfoil surfaces and other surfaces
surfaces.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements
t
The purpose
p p
of load measurements on the model is to make available the forces and
moments so that they may be corrected for tunnel boundary and scale effects and
utilized
tili d in
i predicting
di ti the
th performance
f
off the
th ffull-scale
ll
l vehicle
hi l or other
th device.
d i
The loads may be obtained by at least the following four methods
methods, which are listed in the
order
d off ffrequency off use: (1) measuring
i the
h actuall forces
f
and
d moments on the
h complete
l
model or on parts of the model with one or more balances; (2) measuring the stress
distribution over the model by means of orifices connected to pressure measuring
devices or other means such as pressurep
or shear-sensitive coatings;
g ; (3)
( ) measuring
g the
effect that the model has on the airstream by wake surveys and tunnel wall pressures;
and
d (4) measuring
i the
th motion
ti off the
th model
d l under
d the
th action
ti off the
th aerodynamic
d
i forces
f
and computing the forces from equations of motion.
In this course, we will consider onlyy the first of these, as force balances are universallyy
used in all wind tunnels.
tunnels
AE 2751
July Nov 2016

July-Nov.
24
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
The two most used reference frames are body
y axis frame and wind axis. A third
reference frame is referred to as stability axes.
axes Any reference frame is determined by its
orientation relative to some other frame or a basic physical reference and the location of
th origin.
the
i i A reference
f
frame
f
iis a sett off th
three orthogonal
th
l axes, by
b convention
ti always
l
labeled in a right
right-hand
hand sequence.
Wind
Wi
d Axes:
A
F
For wind
i d tunnel
t
l applications,
li ti
we first
fi t consider
id the
th wind
i d axes. We
W illustrate
ill t t
this in Fig.
g 24. The wind axes have xw p
pointing
g into the wind,, zw p
pointing
g down,, and yw
pointing to the right looking into the wind.
wind If the test section is not horizontal,
horizontal then an
appropriate local convention must be adopted
adopted. We show vectors indicating components
off fforces that are used in wind axes. Note that drag
g is in the negative
g
xw direction and lift
f
is in the negative zw, direction while the side force is in the positive yw direction.
We note
W
t th
thatt a perfectly
f tl aligned
li
d wind
i d tunnel
t
l would
ld have
h
the
th wind
i d axes exactly
tl parallel
ll l to
t
the wind tunnel axis. In reality,
y, there are angularities,
g
, and these lead to the wind axes for
any given model that may be angularly offset from the tunnel axes.
axes
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
Fig 24 Wind and body reference frames

Fig.
AE 2751
July Nov 2016

July-Nov.
25
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
Bodyy axes: The bodyy axe are fixed
f
to the model and move with it. The exact alignment
g
with anyy particular
p
model must be specified
p
as part
p of test planning.
p
g The xb-zb p
plane is
frequently a plane of symmetry or approximately so.
so The force components on body axes
are sometimes referred to as axial force,
force side force,
force and normal force for the xb, yb, zb
components respectively,
components,
respectively or sometimes as body drag,
drag body lift and body side force.
force This
multiple
lti l terminology
t
i l
can lead
l d to
t confusion
f i and
d requires
i
attention
tt ti to
t avoid
id errors in
i
communication. We will use the former set to refer to bodyy axis components.
p
Moments and Reference Frame Origins: The moment components on the xx, yy, z axes
are referred to as rolling moment,
moment pitching moment,
moment and yawing moment,
moment respectively
respectively. If
context
t t iis nott sufficient
ffi i t tto di
distinguish
ti
i h between
b t
what
h t is
i intended,
i t d d then
th body
b d roll,
ll body
b d pitch
it h
and body
y yyaw or wind roll,, wind pitch
p
and wind yaw
y
must be used. Note that the origins
g
of
the reference frames must be carefully specified in every case since the moments are
directly and critically dependent on this choice and there is no universal standard.
standard
Model Attitude: The standard wayy to specify
fy model altitude is to use an Euler angle
g
sequence
q
going
g
g from wind axes to bodyy axes of yaw,
y
, , about the zw axis,, "pitch,
p
, ,
about an intermediate axis
axis, and "roll"
roll , about the xb axis.
axis However,
However it is the aerodynamic
angles angle of attack,
angles,
attack ,
and sideslip,
sideslip ,
that are the preferred independent variables for
writing aerodynamic functions.
functions
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
AE 2751
July Nov 2016

July-Nov.
26
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
AE 2751
July-Nov.
July Nov 2016
27
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
AE 2751
July-Nov.
July Nov 2016
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t R
Reference
f
Frames
F
AE 2751
July Nov 2016

July-Nov.
28
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Moment
M
t Transfers
T
f
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
We have
W
h
been
b
treating
t ti relations
l ti
between
b t
force
f
and
d momentt components
t on different
diff
t
reference frames. A wind tunnel balance is expected to separate these force and
moment components and accurately resolve what is almost always small differences in
large
g forces. A complicating
p
g factor is that the various force and moment components
p
vary widely in value at any given air speed and each varies greatly over the speed
range ffrom minimum
i i
tto maximum.
i
B
Balance
l
d
design
i and
d use are problems
bl
th t should
that
h ld
not be deprecated; in fact, it might truthfully be said that balance design is among the
most trying problems in the field.
field
Concept of a Six
Six-Component
Component Balance: In an attempt to picture the situation most
clearly a conceptual but impractical wire balance based on spring scales is shown in
clearly,
Fig.
g 26. The model,, supposed
pp
to be too heavyy to be raised byy the aerodynamic
y
lift,, is
held by six wires.
wires Six forces are read by scales A,
A B,
B C.
C D,
D E,
E and F.
F The wires attached
t A and
to
d B are parallel
ll l to
t the
th incoming
i
i air
i velocity
l it vector
t and
d define
d fi a plane
l
that
th t can be
b
taken as a reference plane for the balance. We will designate this the xx-y
y plane.
AE 2751
July Nov 2016

July-Nov.
29
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
t Balances
B l
p
These wires point
p
in the x direction. The
wire attached to F is perpendicular to the A
and B wires and is in the x-y
x y plane
plane. This
i points
i t in
i the
th -y di
ti
Th
i
wire
direction.
The wires
attached to C and D are in a plane that is
perpendicular to the xx-yy plane
plane, which we
d i
designate
t the
th y-z plane.
l
Th
The C and
d D wires
i
are p
perpendicular
p
to the x-yy p
plane. Wires A
and C are attached at a common point on
the right wing
wing. Wires B,
B D,
D and F are
f
attached to a common point on the left
wing. Finally the wire attached to E is
parallel to C and D and is in a plane parallel
t C and
to
d 0 and
dh
halfway
lf
between
b t
them:
th
Fig 26 Di
Fig.
Diagrammatic
g
ti wind
i d tunnel
t
l balance
b l
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
1 Since the hori
1.
horizontal
ontal wires
ires A,
A B,
B and F cannot transmit bending,
bending the vertical
ertical
f
force
perpendicular
di l to
t Y
Y, th
the lift,
lift iis obtained
bt i d from
f
the
th sum off the
th forces
f
i the
in
th
vertical wires: L=C+D+E.
2 The drag is the sum of the forces in the two horizontal wires parallel to the
2.
direction of V: D=A + B.
B
3. The side force is simply
p y Y=F.
4 If there is no rolling moment,
4.
moment that is no moment component in the direction
of the x axis
axis, scales C and D will have equal readings.
readings But more generally a
rolling moment will
ill appear as I=(C
I (C - D) X b/2.
b/2 Note carefully
caref ll that this
i with
is
ith reference
f
to
t a point
i t halfway
h lf
between
b t
the
th two
t
wires
i
C and
d D through
th
h
which the line or action of F passes,
p
and in the plane
p
defined as containing
g E.
5 Similarly a yawing moment,
5.
moment that is a moment component in the direction
of the z axis
axis, will result in nonequal forces in the wires A and B and the
yawing moment will be given by n=(A - B) X b/2.
b/2 Here also note that
thi iis a momentt with
this
ith reference
f
to
t a point
i th
halfway
lf
between
b t
A and
d B and
d
through
g which the line off F passes.
6 The pitching moment is given by m=E X cc. This is a moment about the
6.
line containing F.
F
AE 2751
July Nov 2016

July-Nov.
30
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Exactt perpendicularity
E
di l it between
b t
the
th wires
i
mustt be
b maintained.
i t i d For
F instance,
i t
if the
th
wire to scale F is not exactly perpendicular to wires A and B, a component of the
drag will appear (improperly,
(improperly of course) at scale F and will be interpreted as side
f
force.
A similar
i il situation
it ti exists
i t in
i regard
d to
t lift and
d drag
d
and
d lift and
d side
id force.
f
Si
Since
the lift is the largest
g
force by
y far in typical
yp
aircraft complete
p
model wind tunnel work,,
extreme care must be taken to ensure that it is orthogonal to the other components.
components
To illustrate the situation in more detail,
detail consider a planar subset of lift,
lift drag and
pitching moment
moment, as indicated in Fig.
Fig 27.
27 We will assume that we can determine
precisely
i l the
th direction
di ti off th
the x and
d z axes and
d apply
l loads
l d along
l
those
th
axes in
i
order to explore
p
the reaction in the wires. The expressions
p
for equilibrium
q
in the X
and Z directions are
Z+A sin -C
C cos -E=
E= 0
X+A cos -C
C sin =0
m+Ec=0
(11 )
Assume
Ass
me the act
actual
al loads are m=0
0 and Z
Z=10X.
10X What will
ill o
ourr balance read with
ith the
improper
p p alignment
g
as indicated byy Fig.
g 27. Solving
g for A,, C,, and E gives
g
E=0
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
t Balances
B l
p
Fig.
Fig 27 Effect of balance component skew
AE 2751
July Nov 2016

July-Nov.
31
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Four types
yp of external balances have been in general
g
use. These balances are named
for their main load-carrying
y g members-wire,, platform,
p
, yoke,
y
, and pyramidal-and
py
are discussed in the following
g paragraphs.
p g p
Wire Balances: One or the earliest types of wind tunnel balance was the wire balance,
balance
similar in principle to Fig.
Fig 26.
26 Spring scales were not used for the balance output since
their deflections would change the model attitude.
attitude The model usually was mounted
inverted so that aerodynamic lift added to the weight to prevent unloading the wires as
the resulting tension can never be allowed to diminish to zero.
zero With this type of balance
there was a large tare drag on the wires that was difficult to assess accurately.
accurately The
wires
i
tended
t d d to
t break,
b k which
hi h could
ld lead
l d to
t the
th lloss off th
the model.
d l Wi
Wire balances
b l
turned
t
d
outt tto b
be much
h lless robust
b t and
d versatile
til than
th the
th alternatives
lt
ti
and
d have
h
nott been
b
used
d
extensively
t
i l since
i
the
th very early
l days
d
off aeronautics.
ti
Platform,, Yoke,, and Pyramid
y
Balances: Currently,
y, most external balances provide
p
struttype
yp mounting
g of models. These balances provide
p
mechanisms for changing
g g the angle
g
of attack and yaw and transmit the model loads down into a system of linkages that
separate them into force and moment components. Such an apparatus is shown
diagrammatically in Fig.
Fig 28
28, and a linkage system is shown in Fig.
Fig 29.
29 The general
massiveness of a balance structure may be seen in Fig
Fig. 30
30.
AE 2751
July Nov 2016

July-Nov.
32
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Fig. 28 Greatly simplified diagrammatic sketch of balance components
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
t Balances
B l
p
Fig 29 Balance linkage

Fig.
linkage. Lift linkage (not shown) is beneath the roll table
AE 2751
July Nov 2016

July-Nov.
33
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Fig 30 Massiveness of a wind tunnel balance is well illustrated by this photograph of a

Fig.
balance designed for a 150
150-mph
mph wind tunnel with a 9
9-ft-diameter
ft diameter test section.
section
D ring this early
During
earl set
setup
p in the factory
factor the load members have
ha e been dropped in
place
l
without
ith t going
i through
th
h their
th i respective
ti windshield
i d hi ld supportt bases.
b
As
A
shown,
h
th
the balance
b l
has
h approximately
i t l 45O negative
ti yaw
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Tracing
g the p
pathwayy followed byy the loads from model to measuring
g unit (Fig.
( g 28),
) we
see first that the model as illustrated is supported
pp
on two front-load members or "struts"
and a tail strut. The struts,, in turn connect to the inner part
p of a floating
g ring
g frame that is
free to turn (model yaw) and a mechanism is provided to raise or lower the tail strut to
produce model pitch
pitch. The outer pan of the floating frame is held in place by a system of
struts that are specially designed to be strong in tension and compression but very
weak in bending.
bending These struts separate the components of the load by means of a
linkage system and feed them into the measuring unit or output transducers.
transducers Above the
floating frame is a fairing turntable on which are mounted the windshields that minimize
the direct aerodynamic forces on the support struts
struts. The load turntable is tracked by the
f i i turntable
fairing
t t bl through
th
h the
th use off a servomechanism
h i
arrangement.
t And,
A d as the
th fairing
f ii
t t bl rotates.
turntable
t t
Th
The windshields
i d hi ld are rotated
t t d iin the
th opposite
it sense by
b a gear-driven
di
mechanism
h i
so that
th t th
they remain
i parallel
ll l to
t the
th airstream.
i t
In
I some balances
b l
the
th tail-strut
t il t t
fairing
g is moved up
p and down to keep
p the exposed
p
length
g of tail strut constant as the
angle
g of attack is changed.
g
The windshields are connected electricallyy so that upon
p
contact with the load members they
y activate fouling
g lights
g
and/or audible signals
g
so that
the malfunction may be noted and corrected.
The linkage system by which the force and moment components are separated have
gradually worked into three different fundamental types.
types These are named platform,
platform
yoke,
k and
d pyramidal,
id l according
di to
t th
the manner iin which
hi h the
th main
i system
t
is
i assembled.
bl d
AE 2751
July Nov 2016

July-Nov.
34
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Platform Balance: The platform balance (Fig.
(Fig 31) utilizes either three or four legs to
supportt the
th main
i frame.
f
For
F the
th three-legged
th
l
d type,
t
the
th forces
f
and
d moments
t are
L=-(a+b+c)
D=d+e I=2(a-b)(w) n=2(e-d)w S =-f
m=cx
Platform balances are widelyy used,, rugged

gg and naturallyy orthogonal,
g
, theyy mayy be
constructed and aligned with a minimum of difficulty.
difficulty But they also have disadvantages:
(1) the moments appear as small differences in large forces,
forces an inherently undesirable
arrangement; (2) the
h b
balance-resolving
l
l i center is
i not at the
h center off the
h tunnell and
d the
h
pitching moments must be transferred; and (3) the drag and side force loads put
pitching and rolling moments on the load ring.
ring These interactions must be removed from
th fifinall data.
the
d t Some
S
off the
th disadvantages
di d
t
are ameliorated
li t d by
b the
th ease off computation
t ti
today
y as compared
p
to several decades ago.
g Also,, applications
pp
such as automobiles and
surface marine vehicles will be mounted on or near the floor which is near the resolving
center for a yoke balance.
balance
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Fig. 31 Basic layout of a platform balance
AE 2751
July Nov 2016

July-Nov.
35
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Yoke Balance: The yoke balance (Fig.
(Fig 32) offers an advantage over the platform
b l
balance
in
i that
th t the
th moment-resolving
t
l i center
t is
i near the
th center
t off the
th tunnel.
t
l However,
H
the inherent design of the yoke leads to bigger deflections than the platform balance,
Particularly in pitch and side force.
force Because the balance frame must span the test
section
ti iin order
d to
t gett th
the ttwo upper d
drag arms iin their
th i positions,
iti
the
th yaw lever
l
arm is
i
exceptionally
p
y long.
g The high
g supporting
pp
g pillars
p
are subject
j
to large
g deflections.
Once again the final forces must be summed up: The drag is the addition of three
forces and the lift is the sum of two forces in the variant shown
forces,
shown. The yoke balance
as shown
h
here
h
brings
bi
out the
h pitching
i hi moment in
i the
h drag
d
system instead
i
d off in
i the
h
lift. For the yoke balance, the forces and moments are
L ( +b)
L=-(a+b)
D +d+
D=c+d+e
AE 2751
I 2(b )( ) n=2(d-c)w
I=2(b-a)(w)
2(d )
S =ff
m=-ex
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Fig 32 Basic layout of a yoke balance

Fig.
AE 2751
July Nov 2016

July-Nov.
36
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Pyramidal
P
id l Balance:
B l
Th
The complaints
l i t usually
ll heard
h d about
b t the
th platform
l tf
and
d yoke
k balances
b l
are largely
g y overcome byy the ingenious
g
engineering
g
g of the pyramidal
py
type.
yp However,, as
usually happens,
happens additional difficulties are added.
added These are the advantages: The
pyramidal balance reads the moments about the resolving center,
center and the six
components are inherently
i h
l separated
d and
d read
d directly
di
l by
b six
i measuring
i units.
i No
N
components need be added, subtracted, or multiplied. The difficulties involved in reading
the small differences in large forces are eliminated,
eliminated and direct reading of the forces and
moments
t simplifies
i lifi the
th calculations.
l l ti
Note
N t that
th t this
thi is
i less
l
off an advantage
d
t
today
t d than
th it
was several decades ago
g when these systems
y
were being
g intensivelyy developed.
p
Several criticisms of the pyramidal balance are warranted.
warranted The alignment of the
inclined struts is so critical that both the construction and the calibration of the balance
are greatly complicated. Furthermore (and this appears quite serious), deflections of the
inclined struts may so change their alignment that the moments are not accurate.
accurate This
effect
ff t mustt b
be th
thoroughly
hl iinvestigated
ti t d d
during
i th
the calibration
lib ti off th
the b
balance.
l
Th
The manner
in which the pyramidal
py
balance separates
p
the moments is not simple
p and it behooves the
engineer and the student to approach the setup using an elementary truss system.
system
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Consider a truss in which two legs arc jointed (Fig.
(Fig 33).
33) The force D,
D acting through the
pin
i joint
j i t O,
O produces
d
only
l tension
t
i in
i OE and
d compression
i in
i OF.
OF No
N force
f
is
i registered
i t d att
A. However, the force G, not acting through O, produces bending in OE, and OE would
collapse unless the force A=aG/b were present.
present If G and b are known,
known the size of the
f
force
A determines
d t
i
the
th point
i t off action
ti off G.
G IIn thi
this manner, if G were a known
k
drag
d
f
force,
its pitching
p
g moments about the resolving
g center 0 would be determined byy the force A.
Though the previous example illustrates the principle of the pyramidal balance in actual
practice
i a considerable
id bl revision
i i is
i required.
i d In
I order
d to prevent the
h llegs off the
h pyramid
id
from being in the airstream. They are cut off at what would be c in Fig. 33. The truncated
legs are then carefully aligned so that their extensions pass through a common point.
point The
complete
l t setup
t iis illustrated
ill t t d in
i Fig.
Fi 34.
34 The
Th forces
f
and
d moments
t are
Lift=total
Lift
total weight on lowest table
Drag=D
Drag
D
Side force=force C
Pitching
g moment=- PXf
Rolling
g moment=RXf
Yawing
g moment=YXa
AE 2751
July Nov 2016

July-Nov.
37
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
t Balances
B l
p
Fig. 33 Two
Two-dimensional
dimensional schematic showing the principle of the pyramid balance
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
F
Force
M
Measurements:
t Six
Si C
Component
p
t Balances
B l
Fig. 34 Pyramidal or virtual center balance
AE 2751
July Nov 2016

July-Nov.
38
Wind Tunnel
T nnel Measurements
Meas rements
I t
Internal
l Balances
B l
IInternal
t
l balances
b l
are used
d extensively
t
i l for
f complete
l t model
d l workk and
d even more
extensively for measuring loads on parts of models. A six-component
six component internal must
address all the problems faced by a six-component external balance,
balance but it must
accomplish
li h its
it purpose within
ithi a tightly
ti htl specified
ifi d and
d highly
hi hl restricted
t i t d volume
l
and
d
shape.
p Even the best internal balance cannot match the overall dynamic
y
range
g and
component independence of most external balances
balances. But the option of matching the
properties of an internal balance closely to a specific need greatly expands the options
available
il bl when
h planning
l
i an experimental
i
t l investigation.
i
ti ti
The two transducer types that are most widely used for internal balances are strain
gages and
d piezoelectric
i
l t i elements.
l
t Of these
th
two,
t
strain
t i gages far
f outnumber
t
b
piezoelectric devices. The reason for this is that although
p
g piezoelectric
p
devices are
extremely stiff and have outstanding frequency response
response, they do not measure steady
loads well.
well In practice,
practice the term internal balance and internal strain gage balance are
effectively
ff i l synonomous.
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
I t
Internal
l Balances
B l
Basic
B
i Aspects:
A
t There
Th
are ttwo basic
b i types
t
off elements
l
t in
i an internal
i t
l balance:
b l
force
f
elements and moment elements. In any six
six-component
component system there will be three force
units and three moment units
units.
Force-measuring
g elements employ
p y either a cantilever beam or a column arrangement.
g
An eccentric column provides greater sensitivity but also allows more deflection,
deflection as
does a single cantilever.
cantilever The choice might well depend on the particular balance size
and
d arrangement needed
d d ffor a specific
ifi model.
d l Th
The axial
i l force
f
gage"
" shown
h
in
i Fig.
Fi 35a
35
is one of the most common types. This unit can be made very sensitive by sizing the
flexures but since the model is attached to the cage,
flexures,
cage it is subjected to the relatively
l
large
normall forces.
f
One
O might
i ht thus
th expectt (see
(
Fi 35b and
Fig.
d 36) an obvious
b i
interaction
i t
ti
to occur because of the deflection of the flexures. A unit designed
g
at David Taylor
y Model
Basin (DTMB),
(DTMB) shown in Fig.
Fig 35b,
35b has reduced this kind of interaction to a minimum.
minimum All
forces except axial are carried by webs as shear or direct tension.
tension A rod transmits the
axial force
f
to a cantilever beam mounting
g the gages.
g g
The arrangement
g
for
f normal force
f
readout is shown typically in Fig. 37. In this case, the wiring is arranged so that the
difference of the two moments electrically.
electrically
AE 2751
July Nov 2016

July-Nov.
39

I t
Internal
lB
Balances
l
Fig 35 Axial force gage

Fig.
AE 2751
July Nov 2016

July-Nov.

I t
Internal
lB
Balances
l
Fig 36 Normal force interaction on axial force gage

Fig.
AE 2751
July Nov 2016

July-Nov.
40

I t
Internal
lB
Balances
l
Fig. 37 Normal force and pitching moment gage arrangement: (a) differential
circuit for normal force: (b) summing circuit for pitching moment
AE 2751
July Nov 2016

July-Nov.
Wind Tunnel
T nnel Measurements
Meas rements
I t
Internal
l Balances
B l
Since the
Si
th normall force
f
iis equall to
t the
th difference
diff
off the
th two
t
moments
t divided
di id d by
b the
th
distance between gages,
g g
the unit may
y be calibrated directlyy in terms of applied
pp
normal
force.
o ce It is
s important
po a that
a bo
both gage sstations
a o s have
a e the
e sa
same
e sec
section
o p
properties,
ope es, Ily, a
and
d
matched gages
gages. The greater the gage spacing,
spacing the more accurate the normal force
readout If Mf and Mr, are the front and rear moments
readout.
moments, then the normal force N is given by
(Mr - Mf)/d,
)/d where d is the spacing between gages.
gages It is noted that the same arrangement
may be
b used
d to
t measure side
id force.
f
Pitching or yawing moments may be measured by the same gage arrangement discussed
above
b
except,
t as shown
h
in
i Figure
Fi
7.48,
7 48 the
th b
bridge
id is
i connected
t d as a summing
i circuit.
i it The
Th
differential circuit employed
p y for normal force will also yield
y
the moment if the moment
reference point is between the two gage stations, for, then,
Mref=Mf+Xref ((Mr-Mf))/d
where
h
Xref is
i measured
d from
f
th
the ffrontt gage station.
t ti
A
Another
th way iis indicated
i di t d in
i Fig.
Fi 38,
38
where the p
pitch gages
g g are "stacked" and located between the normal force gages.
g g
For
rolling moment a torque tube or a double
double-beam
beam type with gages mounted on the side faces
of the beams can be utilized
utilized. There are many mechanical variations in internal balance
design but the basic arrangement of strain gages attached to flexures is common to all.
design,
all
AE 2751
July Nov 2016

July-Nov.
41

I t
Internal
lB
Balances
l
Fig 38 Three component strain gage balance (NASA design)

Fig.
AE 2751
July Nov 2016

July-Nov.
42
Flow Visualization
Introduction
It is difficult to exaggerate the value of flow visualization. A reasonable mental image
of a flow about a body is almost always necessary for a person to have a useful
understanding of an aerodynamic or hydrodynamic problem. This is true whether the
approach
pp
is strictly
y theoretical,, mainly
y experimental,
p
, mainly
y computational,
p
, or some
combination as is always the most effective. The ability to see flow patterns on and
around a device under investigation often gives insight into a solution to an
aerodynamic problem. This sometimes happens when the pattern of flow exhibited by
the experiment or produced by a computation is in some significant way different from
the mental image that the aerodynamicist had formulated. Or, perhaps, the
aerodynamicist realized there were two or more possibilities and the experimental
evidence resolved the uncertainty. An important reason for the wide appreciation of
computational
t ti
l flfluid
id d
dynamics
i iis th
thatt th
the processes required
i d tto reach
h any solution,
l ti
whether or not it is consistent with physical reality, also strongly support very visual
presentations of very detailed results. These visual representations of detailed data
sets are very memorable, as are many types of flow visualizations in physical
experiments, and both are thereby quite useful for similar reasons.
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
Classic flow visualization for low-speed flows is done by putting something that is
visible into the flow at convenient locations and watching how the something, called a
tracer, moves. The flow is inferred by the motion of the tracer. Tracer methods are the
oldest and most commonly applied means of flow visualization. It seems likely that
every human being has seen naturally occurring smoke or cloud formations being
convected by movement of air and thereby conveying information about the motion of
the air. We must, however, take care to relate how these "pictures" imprinted on our
consciousness are related to our mathematical equations, which we use to quantify
flow phenomena. In addition, it is necessary to investigate to what degree various
sizes and types of tracer particles actually follow the flow.
For the moment, assume that it is possible to put "tracer" particles into the flow at any
desired location and that these particles are then convected perfectly along with the
flo or that we
flow,
e simply
simpl have
ha e the capabilit
capability to "see" an
any small element in the flo
flow that
we choose. We have the following definitions.
Pathline: The path of a point or particle convected with the flow is called a pathline. If
we could release a tracer particle at any selected point and record its subsequent path
this would be a pathline. If we knew the functions indicated by (4), we could construct
pathlines by the parametric relations.
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
AE 2751
July-Nov. 2016
Flow Visualization
Introduction
AE 2751
July-Nov. 2016
Flow Visualization
Direct Flow Visualization Techniques
We consider two broad categories of flow visualization. The first is surface flow
visualization and the second is flow field visualization. Strictly speaking surface
now is also a flow field, but we will use these terms to distinguish between on body and
off-body fields.
fields The methods considered for direct-surface flow visualization include tufts,
tufts
oil now, ink dot, china clay, and liquid crystals. The methods considered for off-body
visualization include smoke injected in several ways, helium bubbles, and streamers.
Methods of Recording Direct Visualizations: There are basically four methods of
recording direct visualizations. The first, historically most important but least permanent,
method is for the engineer to observe with his or her eyes. Because of depth perception,
one can see three-dimensional patterns and one always sees an evolution of the
phenomena,, not jjust the final average
p
g state. However,, there is no direct p
permanent
record in a form that can easily be put into a report or paper. It is possible, however, to
sketch the patterns as they are observed. To do this efficiently, one needs to prepare in
advance a basic drawing of the model on which streamlines or separated regions will be
sketched when the tunnel is running. An advantage of doing sketches is that the mental
process increases the likelihood of forming cognitive maps that capture the flow patterns.
AE 2751
July-Nov. 2016
Flow Visualization
Direct Flow Visualization Techniques
Other common methods of recording the results of flow visualizations are by film, either
still or movie, by video recording using a standard VCR format and by digital recording
of digitized video, either still or movie. These methods produce a two dimensional
projection of a three-dimensional phenomenon.
phenomenon In principle,
principle two or more cameras can
be used to obtain multiple two-dimensional projections that contain sufficient information
to reconstruct the three-dimensional image. In practice this has rarely been achieved
and is not, at the present time, available for routine use. The state of the art in the
needed technology indicates that such capabilities will be available in the near future. A
situation in which such a system would be of great use would be when using smoke or
helium bubbles to trace flow streamlines past a model. The photographic methods while
requiring more time or developing and printing, yield higher resolution. Video has the
advantage of instant replay. The available resolution for video equipment is rapidly
increasing and is adequate for most aerodynamic work today.
AE 2751
July-Nov. 2016
Flow Visualization
Surface Flow Visualization Techniques: Tufts
Information about the flow on the surface of an object being studied is usually most
critical. Many times, the flow off the body is of interest primarily in order to understand
the flow features on the surface. Key aspects of surface flows that may be investigated
using visualization techniques include stagnation point location
location, separation lines
lines,
location of boundary layer transition, characteristic unsteadiness, extent of separation
zones, and types of critical points and their locations.
The simplest and most frequently used method for surface flow visualization is to attach
tufts to the surface of interest. The tufts must be of light, flexible material that will align
itself with the local surface flow as a result of direct aerodynamic force. The most
commonly used material is light yarn with weights and lengths chosen according to
model size and test speeds.
p
Very
y small monofilament has also been used. There are
also polyester and cotton sewing threads, such as Clark's O.N.T. mercerized cotton No.
60, which can be treated with a fluorescent material. The thread is a multiple-strand
material and tends to unravel with time. Tufts do affect the aerodynamic forces to some
extent as we will show, but there are many situations in which the method is so easy
and economical that it is the first choice.
AE 2751
July-Nov. 2016
Flow Visualization
Two basic methods of attaching tufts to a surface are by scotch tape or by glue. When
tape is used, the tufts are usually made on a "tuft board." The tuft material is strung
back and forth around pins, then the tape is applied to the tufts and the tuft material is
cut at the edge
g of the tape.
p This g
gives a length
g of tape
p with tuft attached that is applied
pp
to the model (Fig.1). The model surface is cleaned with naphtha or other solvents to
remove oil so that the tape will hold under the adverse conditions of high-speed flow.
When tufts are glued to the model, a nitrocellulose cement such as Duco is used,
thinned 50% with acetone or methyl ethyl ketone. Often 10% pigmented lacquer is
added both to obscure the portion of the tuft under the glue and to make the glue dots
visible by using a contrasting color. The glue dots are kept as small as possible. Tufts
readily show where flow is steady and where it is unsteady. Regions of complete
separation and buffeting flow are readily identified. The resolution of the determination
is of the order of the spacing of the tufts. The possibility of significant influence of the
tufts themselves on the flow is very high and must always be kept in mind. This can be
investigated by removing tufts upstream of indicated flow separation.
An example using No.6 floss (crochet yarn) is shown in Fig. 2. An example using No.
60 thread tufts is shown in Fig. 3 for a transport aircraft wing.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 1 Methods of taping tufts to model and a tuft board. The pattern shown as A is used for high-speed tests
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 2 No. 6 floss (crochet yarn), white light source: =27.3O. Tufts taped to wing as in method B of Fig. 1.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 3 No. 60 thread tufts glued to wing, UV source: =27.3O
AE 2751
July-Nov. 2016
Flow Visualization
Surface Flow Visualization Techniques: Minitufts
The glue technique is used for minitufts. These tufts have the least effect to the
aerodynamic data and thus are often left on the model. The tuft material is mono
filament nylon that has been treated with a fluorescent dye. Two sizes are used: 3
denier (diameter 0.02 mm, 0.0007 in.) and 15 denier (diameter 0.04 mm, 0.0017 in.).
The dye used is Leucophor EFR liquid in a concentration of 1% in water with 2% acetic
acid added. The tuft material is wound on an open wire reel and immersed in the dye
for 15 min at 82.2C (180F) with frequent agitation. After drying for at least 1 hr the tuft
material is wound onto small spools. During this step the material should be wiped with
tissue pads to remove loose fluorescent powder that can transfer to the model surface
in irregular patterns. When gluing tufts to a model, a square grid is used (typically
about 0.75 X 0.75 in.). The tuft material is taped to the wing undersurface and then
wrapped around the wing in a chordwise direction
direction. The material is in the chordwise
direction on the top and moves diagonally across the wing on the bottom surface. As
an alternate, the tuft material can be taped at both the leading and trailing edges. After
the tuft material is applied, it is glued using a hypodermic syringe with a fine needle (a
coarse needle can be partially closed with pliers). As the desired size drop of glue
forms on the needle, touch it to the surface and pull away quickly.
AE 2751
July-Nov. 2016
Flow Visualization
Surface Flow Visualization Techniques: Minitufts
After the glue dries the tufts are cut just ahead
of the glue spot of the next tuft. The model
surface should be cleaned before the tufts
and g
glue are applied
pp
using
g Freon or
chlorinated hydrocarbons. The monofilament
nylon minitufts acquire static charges. These
can be neutralized by the use of antistatic
solutions or the antistatic material used in
home dryers. The minitufts are viewed and
photographed in ultraviolet light as this is the
way to make the small monofilament most
visible. Threads can be similarly treated and
can be photographed in either ultraviolet or
white light. Minitufts provide the same type of
information as larger tufts. They can provide Fig. 4 Fluorescent minitufts, UV source: =27.3O.
greater resolution and have less influence on
Wing is marked with a fluorescent felt marker pen
the flow. An example of surface visualization
Compare stalled region near tip with Figs. 3, 5 & 8.
using fluorescent minitufts is shown in Fig. 4.
AE 2751
July-Nov. 2016
Flow Visualization
Surface Flow Visualization Techniques: Oil
Oils and other viscous fluids are used to show the surface flow. The selected material
is usually spread on the areas of interest with a paint brush. It will then flow under
the influence of shear stress from the air stream and gravity. Since inclined surfaces
are almost always of interest, the mixture needs to have viscosity sufficient so that
it will not flow rapidly under the influence of gravity. The flow speed of the air
must then be sufficient to impress shear stress large enough to cause the oil to flow
and reveal the surface patterns within an acceptable time. This is typically of the
order of 10's of seconds after the tunnel is brought to speed. It is difficult to use
oil flow on vertical surfaces at air speeds less than 100 mph and 150 mph is much
better. The most common material for oil flow is petroleum lubricating oils. These
materials are messy to clean up afterward, both on the model and more so in the
t
tunnel.
l Another
A th material
t i l that
th t works
k as wellll as oilil when
h
t t d with
treated
ith a fluorescent
fl
t
dye is polyglycol. At high CL's or high surface velocities this material may have
too low viscosity, making it difficult to use. This material can be cleaned up with
soap and hot water. When cleaning the tunnel after extensive oil flow runs, a portable
set of ultraviolet fluorescent tubes is most useful. If the oil flows too slowly, it is thinned
with naphtha, and if it is too thin, 60-70W oil is added.
AE 2751
July-Nov. 2016
Flow Visualization
The viscosity of the mixture is adjusted by trial and error for each application. The color of
the oil needs to contrast with the color of the model surface. A widely used method is to
add a fluorescent dye to the oil and illuminate it with ultraviolet lights, as is done for the
fluorescent mini tufts. In this case the model color is not too critical so long as it does not
reflect strongly under the ultraviolet illumination.
illumination A light blue works very well with a dye
that provides fluorescence in the yellow region of the spectrum. This is the most
commonly used combination. An example result of oil flow visualization using 40W motor
oil treated with a very small amount of fluorescent dye is shown in Fig. 5. Oil can be
made white by mixing titanium dioxide into it. This can be applied to a black model and
ordinary light used for viewing and photography. This is sometimes preferable to the
installation of black lights and the subsequent requirements on light management. Two
examples are presented. Figure 6 shows a low aspect-ratio wing at high angle of attack.
Figure 7 shows the upper surface of an automobile. Photographs can be taken after the
tunnel is turned off,
off but the available time is short even on horizontal surfaces as the oil
will flow under gravity. China Clay China clay is a suspension of kaolin in kerosene. The
fluid is applied with a paint brush, usually with the model set at the desired attitude. The
tunnel is started as quickly as possible after the model L painted. When the mixture has
dried, photographs can be taken after the tunnel is shut down because the pattern does
not change rapidly with time. This is similar to the Fales method with the kaolin
substituted for lamp black. An example of visualization using china clay is shown in Fig. 8.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 5 Oil applied at =0O, tunnel started and brought

close to speed, model pitched to =27.3O
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 6 Oil flow on a low aspect ratio rectangular wing (Clark Y-14 wing, aspect ratio 4, =25.4O, Re=360,000)
AE 2751
July-Nov. 2016
10
Flow Visualization
Fig. 7 Oil flow on an automobile, yaw 0O
AE 2751
July-Nov. 2016
Flow Visualization
Surface Flow Visualization Techniques: China Clay Technique
Fig. 8 China clay applied at =0O, tunnel started and brought close to speed, model pitched to =27.3O
AE 2751
July-Nov. 2016
11
Flow Visualization
Effects of Tufts, Minitufts, China Clay, and Oil
Tufts can affect the aerodynamic loads on a model.
In Fig. 9 a lift curve near stall shows the effects of
various tufts on the data. The g
glued minitufts and
No. 60 thread tufts consisted of about 900 tufts. The
two taped tufts consisted of about 300 tufts. The
data are an average of five runs for each set of tufts.
The mini tufts and the glued No. 60 thread have the
minimum effect on lift. The effect of the tape can be
seen by comparing the two sets of No. 60 thread
tufts. The six-strand floss tufts are similar to the tufts
made out of yarn. The three different tuft types can
b seen att =27.3
be
27 3O in
i Figs.
Fi
24
2-4.
Fig. 9 Effect of various tufts, china clay and oil

flow on lift curve near stall
AE 2751
July-Nov. 2016
Flow Visualization
Sequencing Attitude and Speed Setting
The usual procedure for china clay
visualization is to set the model at the
desired angle of attack and bring the
tunnel up to speed. On a wing with a
slotted leading-edge and/or trailingedge flaps. This can result in erroneous
aerodynamic data and flow visualization
due to flow separation in the slots at
low Reynolds numbers during the
tunnel acceleration. This is shown by
the oil flow and china clay data points
y
=27.3O in Fig.
g 9. Similar data
beyond
were obtained on the clean wing.
Figure 10 shows a china clay flow
visualization for this test method.
AE 2751
Fig. 10 China clay applied at a=27.3O, tunnel started and

brought to close to speed. Leading edge slotted
flap is stalled. Compare with Figs. 4 and 8
July-Nov. 2016
12
Flow Visualization
Symmetry and Hysteresis
An example of a very interesting flow phenomenon is given in Figs. 11 and 5.12. Although
it is common to assume that symmetric boundary conditions produce symmetric flows,
there are many counter examples. Any flow that includes large regions of separation may
well exhibit asymmetry of the mean flow as well as asymmetry of the instantaneous flow
even if the solid boundaries are sensibly symmetric. This can lead to results such as that
illustrated here in which the integrated forces and moments exhibit random switching or it
can lead to hysteresis in which the forces and moments will be dependent on the time
history of the altitude. Flow visualization methods are often important tools in identifying
the flow structures associated with such events.
Flow Topology: Topological concepts are slowly gaining recognition as important to the
study of complex flow phenomena such as bluff body flows and flows about wings and
aircraft at high angles of attack. The concepts are providing ways to understand the
structure of flow fields and to tie the structure of a flow field to the topology of the surface
flow on the test article. As oil flow techniques provide very fine detail of the surface flow,
this is a preferred technique for studies of surface flow topology.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 11 Oil visualization of flow on an airplane at =14O, that has

flow on right wing attached and on the left wing separated
AE 2751
July-Nov. 2016
13
Flow Visualization
Fig. 12 Here the flow on the left wing attached and on the right wing separated. The model is
not moved. The flow state alternated between the two conditions shown at random
intervals in the range of a few seconds at a chord Reynolds number of about 2x106.
This produces a very strong rolling moment that changes very abruptly and randomly.
AE 2751
July-Nov. 2016
Flow Visualization
Some Procedural Details
Correlation of Balance Data and Flow Visualization: The generally accepted practice
when using the older, larger, six-strand floss for tufts is not to take balance
measurements, or at least not to consider them to produce good data when tufts are
applied to the model
model. Figure 9 explains why this is the practice
practice. Both the lift curve slope
and maximum lift are greatly reduced. An advantage of minitufts is that their effect on
the data is minimal; hence they can be left on the model. Oil and china clay also show
minimum aerodynamic effects. During this comparison test of tufts at the model's
minimum drag (a lift coefficient of about 1.0). The model drag decreased as tufts from
the mini to No.6 floss were added to the left wing with the horizontal tail on and the
reverse happened with the tail off. The tufts apparently change the wing's span load
distribution. However, old practices die slowly, so force data are often not taken during
surface flow visualization. This, then, can lead to an improperly established flow field
and the possibility of misleading now visualization, especially near stall and, oddly
enough, near minimum drag. If, however, force data are taken before and during the
flow visualization run, the error may be detected and thereby possibly avoided.
AE 2751
July-Nov. 2016
14
Flow Visualization
Effects of Tufts, Minitufts, China Clay, and Oil
Tufts can affect the aerodynamic loads on a model.
In Fig. 9 a lift curve near stall shows the effects of
various tufts on the data. The g
glued minitufts and
No. 60 thread tufts consisted of about 900 tufts. The
two taped tufts consisted of about 300 tufts. The
data are an average of five runs for each set of tufts.
The mini tufts and the glued No. 60 thread have the
minimum effect on lift. The effect of the tape can be
seen by comparing the two sets of No. 60 thread
tufts. The six-strand floss tufts are similar to the tufts
made out of yarn. The three different tuft types can
b seen att =27.3
be
27 3O in
i Figs.
Fi
24
2-4.
Fig. 9 Effect of various tufts, china clay and oil

flow on lift curve near stall
AE 2751
July-Nov. 2016
Flow Visualization
Relative Advantages of Tufts
Tufts have a great advantage in terms of productivity. Once the tufts are installed, the model
can be repositioned and the indications studied visually and photographed for as long as
desired. The model can then be simply
p y moved to a new condition and the p
process continued.
An example of a series of tuft photographs for a sequence of conditions for a powered tilt
wing half model is shown in Fig. 13.
The oil and china clay method produce patterns of limited duration in time for a single
operating condition. Then the tunnel must be stopped and the material reapplied for each new
condition to be visualized. Tufts are sometimes used for broad looks over a wide range of
conditions with the more detailed techniques then applied for a smaller and selected set of
conditions that have been found to be most critical.
AE 2751
July-Nov. 2016
15
Flow Visualization
Relative Advantages of Tufts
Fig. 13 Visualization of flow on a powered wing model in a series of conditions using tufts
AE 2751
July-Nov. 2016
Flow Visualization
Boundary Layers and Surface Shear Stress
Often the most important information being sought by flow visualization methods is a
definition of the locations of transition from laminar to turbulent boundary layers and the
locations of any separation regions. In the previous section, we have shown several results
that illustrate separated flows
flows. The location of transition between laminar and turbulent now
cannot be determined by tufts and is difficult with china clay. However, oil flow, sublimation
methods, infrared images and liquid crystals can be used to locate transition.
Oil Flow Detection of Transition: The basis for detecting boundary layer transition by viewing
oil flow patterns is the increase in wall shear stress when a boundary layer transitions from
laminar to turbulent. The result is that the oil is swept away faster in the region where the
boundary layer is turbulent. Transition indicated by oil flow visualization is shown in Fig. 14.
The shear stress at the leading edge of a surface is high even under a laminar boundary layer
so a typical
t i l pattern
tt
is
i th
thatt the
th oilil iis sweptt away rapidly
idl att th
the lleading
di edge
d with
ith a gradual
d l
lessening of the scrubbing as the laminar layer thickens and then severe scrubbing
downstream of transition. The oil pattern given by a laminar separation bubble with turbulent
reattachment can be seen just downstream of the leading edge of the wing in Fig. 6. In some
cases a subcritical bump in the oil can cause a wake, which can confuse the transition
location. This does not occur with sublimation.
AE 2751
July-Nov. 2016
16
Flow Visualization
Fig. 14 Oil flow visualization showing natural transition at approximately

40% chord; CL=0.28, Re=1,26,000 based on average chord
AE 2751
July-Nov. 2016
Flow Visualization
Sublimation: In one sublimation technique a mixture of naphthalene and a carrier such as
fluorine, acetone, or methyl ethyl ketone is sprayed on the model using a standard air
spray gun. Note that the last two can remove many paints. The operator must wear a
respirator
p
mask when doing
g this. The mixture will leave the model surface white and
therefore works best on a black or dark surface. The turbulent boundary layer will scrub the
mixture off. A natural transition is shown on a wing using naphthalene in Fig. 15.
Liquid Crystals: Liquid crystals that undergo changes in reflective properties as
they are exposed to shear stress can be used as detectors of transition. This method
has recently been developed as a quantitative method as well as a method of visualization.
Infrared Thermography: The basis for this technique is that a surface at a temperature
different from the tunnel stream will have faster heat transfer from the region under a
turbulent boundary layer than from the region under a laminar layer.
layer The
resulting small temperature differences can be observed using commercially available
infrared cameras. The results will vary with model structure and heat transfer
characteristics so it is advisable to check results for a particular model type by
observing artificially tripped layers. It should be kept in mind that it is the relative
temperature, not the specific value of temperature that gives the indication of transition.
AE 2751
July-Nov. 2016
17
Flow Visualization
Fig. 15 Naphthalene-fluorine flow visualization showing natural transition at

approximately 40% chord; CL=0.28, Re=1,26,000 based on average chord
AE 2751
July-Nov. 2016
Flow Visualization
Ultraviolet Fluorescence Photography: Ultraviolet fluorescence photography is used when
the medium used for visualizing the flow has been treated with a dye that radiates in the
visual spectrum when excited by ultraviolet light. The two common cases already
illustrated are the fluorescent minitufts and fluorescent oil
oil.
The wavelength of the ultraviolet light is 320-400 mm, and it is transmitted by optical
glass. There are three sources that are typically used to produce the ultraviolet light. They
are special fluorescent tubes (black lights), mercury vapor lamps, and photo flash units. In
wind tunnel use, the first two light sources are used enable the test engineer to observe
the flow, and the flash units are used to take still photographs of the flow when desired.
Fluorescent tubes and mercury vapor lamps in general do not have a high enough light
intensity to allow photographs without a very long exposure. Because fluorescent material
emits light in the range, the tunnel test section must be shielded from visible light. Both
mercury vapor lamps and the flash units also emit visible light. Thus they must be
equipped with an exciter filter that will transmit ultraviolet light and absorb visible light.
Glass filters that accomplish this are Kodak Wrattan filter No. 18A or Corning No. 5840.
AE 2751
July-Nov. 2016
18
Flow Visualization
For larger tunnels the flash lamps are usually studio units marketed for commercial
photographers. The flash lamps should be able to handle 2000 W/s per flash as rated
by photographers; the units come with power supplies that can store energy in
capacitors and have the necessary trigger circuits
circuits. The reflectors for the flash units
should be 10-14 in. diameter to be efficient. The Corning glass exciter filter comes in
6.5-in. squares, four of which can be glued together and built into a frame to cover the
reflector. As an alternative, one glass filter can be used with the flashbulb without a
reflector. This would require approximately one additional f stop. For research tunnels
where the camera-to-subject distances are small, standard flash units and a Wrattan
18A exciter, which can be obtained in a 3.5-in. square can be used.
Photographic film is sensitive to blue and ultraviolet light. The light reaching the
camera will contain both the visible fluorescent radiation and reflected ultraviolet
radiation. To prevent the ultraviolet from exposing the film, a barrier filler is attached to
the camera lens. The barrier filter can be a Kodak Wrattan filter No. 2A, 2B, 2C, or 2E.
These can also be obtained in 3.0-in.-square gelatin sheets.
AE 2751
July-Nov. 2016
Flow Visualization
Focusing the camera with a fluorescent light source is usually no problem because this
can be done using either the black-light fluorescent lights or the mercury vapor lamp
with an exciter filter as a light source. It should also be noted that black lights that can
be fitted to standard fixtures are several orders of magnitude cheaper than the mercury
vapor light with its power supply and exciter filler. It is also possible to photograph the
fluorescent material with some video cameras during the flash from the light source.
AE 2751
July-Nov. 2016
19
Flow Visualization
Flow Field Visualization
Tuft Wands and Tuft Grids: The least expensive and at the same time a very versatile
method is using a tuft wand, a long tuft on a pole useful for tracing now near the test
article. If it is necessary that a person be in the runnel. He or she must wear goggles to
protect the eyes from dust
dust. The person should disturb the flow in the vicinity of the test
article as little as possible. Fig. 16 shows a tuft wand in use. To make the flow pattern
downstream of the test article visible, a tufted wire grid is useful, as shown in Fig. 17.
Helium-Filled Soap Bubbles: To trace pathlines, which are also streamlines if the flow
is steady, helium-y filled bubbles that have neutral buoyancy can be used. The bubbles
are inserted ahead of the model and are photographed with a high-intensity light that
passes through
g the tunnel as a plane of light.
g With proper photographic
g
exposure time
the bubbles appear as streaks. Maximum tunnel speed for use of helium bubbles is
about 60 mph (30 m/s). The bubble generators are available commercially.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 16 Investigating the flow structure of an automobile with a tuft wand
AE 2751
July-Nov. 2016
20
Flow Visualization
Fig. 17 Flow visualization by the grid and tuft method, yawed delta wing
AE 2751
July-Nov. 2016
Flow Visualization
Flow Field Visualization: Smoke
Methods of Smoke Production: The most common method of flow field visualization
after tufts is smoke, which can be produced in a number of ways, although a universally
accepted "best" way has yet to be devised. Burning damp straw, rotten wood and
tobacco to produce smoke is generally unsatisfactory, even though historic work was
accomplished with smoke from such materials. The same is true of pyrotechnic smoke
devices (smoke bombs). Chemical methods of producing smoke include both titanium
tetrachloride and tin tetrachloride which produce smoke when brought into contact with
moist air. These compounds produce the very best dense smoke filaments but the
products are corrosive and can be used only in an open circuit facility that can be well
vented to an appropriate location. A mixture of anhydrous ammonia and sulfur dioxide
produces dense white smoke, odors, and, if the air is damp, sulfuric acid. A significant
advantage
g of the system
y
is the fact that the smoke can be started and stopped
pp
in a
fraction of a second, which is not true of the alternatives. This allows the problems
associated with it to be minimized if photographic records rather than real-rime human
visual inspection can be paramount. Steam and liquid nitrogen produce a good dense
smoke with no aftereffects but are very difficult to control and therefore seldom used.
AE 2751
July-Nov. 2016
21
Flow Visualization
A method reported by Shindo and Brask that works at velocities of over 100 mph is a
vaporized petroleum product called Type 1962 Fog Juice, which is used in theatrical
productions. The smoke generator consists of about 75 in. of 0.060-in. outside-diameter
(wall = 0.01 in.) stainless steel tubing. For a probe the tubing is placed inside a 0.375in diameter steel tube and held by a collet about 8.0
in.-diameter
8 0 in.
in from the end of the tubing,
tubing
which is bent 90 about 3 in. from the end. Ceramic beads are used to insulate the
stainless steel tubing from the outer tube. To vaporize the fluid, 10-15 A is applied to the
stainless steel tubing from the collet to a point about 60.0 in. away inside a non-heatconducting handle. This allows the stainless steel tube to expand. The power unit
consists of a variac whose output is connected to the 230-V windings of a 1.5-kVA
115:230 transformer. The variac is used to control the temperature on the stainless
steel tubing by applying 0-50 V. The tubing has about 3 resistance, so the
current is limited to a maximum of 15 A.
A
AE 2751
July-Nov. 2016
Flow Visualization
The fuel reservoir is airtight and has a pressure regulator used to set plant air
pressure at about 30 lb/in2 to feed the fuel to the probe and a needle valve to control
the fuel flow. Plastic tubing connects the reservoir to the probe. It takes some
experience to obtain the desired volume of smoke which is affected by the air
pressure fuel flow and voltage.
pressure,
voltage The following values arc approximate.
approximate With the
power switch off, set the variac at 70%, apply 30 lb/in2 to the reservoir, and crack
open the needle valve. When a small stream of fluid comes out the end, turn on
the power. When turning off, cut the power, and when a stream of fluid leaves the
tube, shut the needle valve off. This is done to reduce carbon formation in the
stainless steel tubing. If hot fluid is emitted, the temperature is too low; either
increase the voltage or reduce the fuel flow. If the smoke pulsates, it is generally a
sign that the air pressure is too low.
The most widely used method for producing smoke at the larger wind tunnels today is
a wand system similar to that described above with polyethylene glycol as the "fuel."
A useful alternative to air pressure to drive the fuel through the wand is a peristaltic
pump. These are sold by suppliers to chemistry and life science laboratories and
have very tine volume flow control. An example of such a system is shown in Fig. 18.
AE 2751
July-Nov. 2016
22
Flow Visualization
The fuel reservoir is airtight and has a pressure regulator used to
set plant air pressure at about 30 lb/in2 to feed the fuel to the probe
and a needle valve to control the fuel flow. Plastic tubing connects
the reservoir to the probe. It takes some experience to obtain the
desired volume of smoke which is affected by the air pressure, fuel
flow and voltage. The following values are approximate. With the
power switch off, set the variac at 70%, apply 30 lb/in2 to the
reservoir, and crack open the needle valve. When a small stream of
fluid comes out the end, turn on the power. When turning off, cut
the power, and when a stream of fluid leaves the tube, shut the
needle valve off. This is done to reduce carbon formation in the
stainless steel tubing. If hot fluid is emitted, the temperature is too
low; either increase the voltage or reduce the fuel flow. If the smoke
pulsates, it is generally a sign that the air pressure is too low.
The most widely used method for producing smoke at the larger
wind tunnels today is a wand system similar to that described
above with polyethylene glycol as the "fuel." A useful alternative to
air pressure to drive the fuel through the wand is a peristaltic pump.
These are sold by suppliers to chemistry and life science
laboratories and have very tine volume flow control. An example of
such a system is shown in Fig. 18.
AE 2751
Fig. 18 Smoke generator and delivery

wand used at the Glenn L. Martin
wind tunnel, University of Marylan
July-Nov. 2016
Flow Visualization
F N M Brown developed a system of smoke generation in which kerosene is
vaporized and the vapor then cooled to the temperature of the airstream before being
emitted into the stream. The systems are designed to be used with special purpose
flow visualization wind tunnels rather than in larger general-purpose wind tunnels.
They have been used in many studies.
studies
Using Smoke: Smoke filaments can be used to find key locations such as stagnation
points. Smoke can easily show the extent of separated regions and the size of a
separation bubble. Smoke can also be used to show and track strong features such as
tip or leading-edge vortices from lifting surfaces that may impinge on downstream parts
of a vehicle with deleterious effects. Strong stable light sources are required. If a
periodic phenomenon is being investigated, then a stroboscope will be needed that can
be synchronized to the period of the phenomenon being studied. Common subjects
requiring such an arrangement are propellers and rotors.
rotors Regions of separated flow can
be detected in two opposite ways. Smoke filaments can be introduced upstream of the
model and located in a series of positions so that the boundary of essentially
undisturbed flow can be detected. Any volume into which smoke does not go readily is
likely a wake region. The smoke wand can then be introduced directly into the apparent
wake region. The smoke will fill the wake "bubble," thereby defining its extent.
AE 2751
July-Nov. 2016
23
Flow Visualization
Flow Field Visualization: Smoke Wire
Smoke Wire: Another method of producing small discrete filaments of smoke at
low velocities is described by Batill and Mueller. This is the smoke wire technique, and it
appears to be limited to flows where the Reynolds number, based on wire diameter,
does not exceed 20 or at velocities from 6 to 18 ft/s (2 to 6 ft/s). The limit is based on
preventing the wake from the wire from disturbing the flow behind the wire,
wire and the limit
has been determined by experiment. This method uses a small-diameter wire that is
coated with an oil. The best results were obtained using Life-Like model train smoke,
which consists of a commercial-grade mineral oil with small amounts of oil of anise and
blue dye added. The liquid-coated wire has 40-80 V AC or 40-60 V DC impressed
across it. As the wire is heated, fine smoke streaklines form at droplets on the wire
(approximately 8 lines/cm for a 0.003-in.-diameter wire). As the wire is heated. It
expands and thus sags. This can cause problems with the accurate placement of the
streaklines To avoid this,
streaklines.
this the wire as prestressed to about 1.5X10
1 5X105 lb/in2, which is near
the yield point of type 302 stainless steel wire; thus the wire must be handled carefully.
As the oil-treated wire produces smoke for periods of up to 2 sec. a timing circuit was
used for the lights and cameras used to photograph the smoke. Since the smoke wire is
limited to low-Reynolds-number tests, it is probably best suited to use in Small tunnels
as it is difficult to run large tunnels at the required low velocities. An example of a smoke
wire visualization is shown in Fig. 19.
AE 2751
July-Nov. 2016
Flow Visualization
Flow Field Visualization: Smoke Wire
Fig. 19 Streamlines past an airfoil visualized by smoke wire technique
AE 2751
July-Nov. 2016
24
Flow Visualization
Data Driven Visualization
All the techniques in which information is processed by either analog or digital methods
and produce flow images of some sort are placed in this category. The data may be a
set of measurements such as the voltages from a set of pressure transducers attached
to multihole probes along with the geometric data that give locations corresponding to
each set of voltage data. Many other types of measurements are used in similar ways.
Another example would be the luminosity data from a CCD camera recording the image
of a model coated with pressure- or temperature-sensitive paint. The data may also
come from a set of computations based on some flow model such as potential flow,
Euler codes, or Navier-Stokes solvers. The demarcation between the class of methods
addressed here and those that have been termed "direct visualization" is rapidly blurring
as every image shown earlier already exists in digitized form and can be manipulated
as a set of numerical data. The evolution is being driven by the continuing advance in
our capability to acquire, manipulate, and present in various forms larger and larger
data sets in smaller and smaller times.
AE 2751
July-Nov. 2016
Flow Visualization
Optical Methods of Flow Visualization
Optical methods of flow visualization can be divided into the following methods:
1.
2
2.
3.
4.
5.
6.
Shadow method
Schlieren method (parallel or focused,
focused gray or color)
Interferometry (classical, holographic)
Electronic speckle interferometry and stereography
Laser Doppler anemometry
Particle Image Velocimetry
The optical methods are mainly used for visualization of compressible flows.
flows
Reference: Ristic Slavica, A view in the invisible, Theoretical and Applied Mechanics,
TEOPM7, Vol. 40, No.1, pp. 87-119, Belgrade, Yugoslavia, 2013.
AE 2751
July-Nov. 2016
25
Flow Visualization
Optical Methods of Flow Visualization
Airflow around aerodynamically models, in optical sense, is a transparent

environment with complex light refraction index. Light refraction index n in each point is
the function of air density, which, on the other side, is the function of speed, pressure
and air temperature. The relation between air density (x,y,z) and refraction
index n(x,y,z) is the Gladstone
Gladstone-Dale:
Dale: n=1+K
1 K. The constant K has dimension of -1
and it is different for each gas.
According to Snell's law, a light ray, passing through inhomogeneous refracted field, is
deflected from its original direction and a light path is different from that of undisturbed
ray. If recording plane is placed in front of light ray, after disturbing media, three
quantities can be measure: the vertical displacement of disturbed ray, the angular
deflection of disturbed ray with respect to the undisturbed, the retardation of deflected
ray, i.e. the phase shift between both rays [1].
Optical visualization methods are based on the recording one of these three
quantities, or a combination of them. Shadowgraph used the first phenomenon, the
Schlieren the second, and interferometry the last. The shadowgraph is sensitive to
changes of the second derivative of density or refractive index 2n/y2, Schlieren to
changes of density or first derivative n/y, and the interferometry is capable to measure
absolute density n changes. If, using the optical method, light refraction index n(x,y,z) in
flow is determined, another physical parameter of tested environment, significant to
aerodynamic testing, can be indirectly determined as well.
AE 2751
July-Nov. 2016
Flow Visualization
Optical Methods of Flow Visualization: Shadowgraph Method
The oldest and the simplest of all optical methods for flow visualization is shadowgraph.
Figure 20a shows the bow shock wave ahead of sphere in supersonic wind tunnel
T-36 at M=1.86. The trace of the shock wave on the photo is a band of absolute
darkness bounded on the downstream side by an edge of intense brightness.
brightness The exact
geometrical position of the shock front is the other edge of the dark zone. The shock
wave represents a jump of the refractive index. The air density increases after the
shock and the incident ray deviates to inside edge. Since the density in the disturbance
is lower than in the surrounding field, (Prandtl-Meyer expansion fan at the sharp end of
the nozzle) the bright band appears at the beginning of the shadow. The same result is
obtained when the compressible boundary layers is visualized. Figure 20b is typical
shadowgraph
g p showing
g flow around a spherical
p
tipped
pp cylinder
y
mounted on flat p
plate [[5].
]
Shadowgraph methods with short duration light pulses can be used for fine visualization
of turbulent compressible flow.
AE 2751
July-Nov. 2016
26
Flow Visualization
Optical Methods of Flow Visualization: Shadowgraph Method
Fig. 20 Shadowgraph visualization of (a) flow around a sphere and (b) typical shadowgraph
images showing spherical tipped cylinder mounted on a flat plate
AE 2751
July-Nov. 2016
Flow Visualization
Optical Methods of Flow Visualization: Schlieren System
As mentioned before, Schlieren method is sensitive to change of the first derivative
of density n/y, (or refractive index) and it can record the angular deflection of the
disturbed ray with respect to the undisturbed in a transparent medium with local
inhomogeneities.
The Schlieren method is the most frequently used in aerodynamic laboratories,
laboratories
since it is relative simple and very useful method. If a parallel beam of light passes
through air with density gradient normal to the direction of the beam, the beam is
refracted towards the region of greater density.
The most simply is the Schlieren system with parallel light through the test section
of the wind tunnel. Tepler system in hypersonic wind tunnel, as the base of all other
systems, is illustrated in Fig. 21.
The modern Schlieren system uses color filter or phase optical elements instead of
the knife-edge,
g and have several p
parallel, transparent,
p
colored strips
p ((most often three
colored sheets, red-blue-yellow or blue-green-red). The color filter can be consisted of
four differently colored strips arranged in a square filter for visualize the grad n in two
directions. If the flow is axis symmetric, complementary colors appear for the same
event (compression or expansion) above and below the flow axis. The recorded pure
colors and color combinations are a measure for the local direction of density gradient
in the test section. Figure 21 shows components of Schlieren systems in T-34
hypersonic wind tunnels in MTI.
AE 2751
July-Nov. 2016
27
Flow Visualization
Fig. 21 Schlieren system components, models in the test section of a

hypersonic wind tunnel and TV camera with monitor
AE 2751
July-Nov. 2016
Flow Visualization
Figure 22 shows black and white color Schlieren images around a bullet at transonic
Mach numbers and muzzle blast from a 22-caliber rifle.
Attempts to increase the amount of information extractable from Schlieren images,
various opaque filters with different geometries, as well as transparent phase and
color filters are used.
Figure 23 shows color Schlieren images around a blunt body and thin protruding probe
mounted in front of a blunt body, used to reduce the drag and the rate of heat transfer,
for M=1.86.
Flow visualization in two dimensional model of the supersonic rocket nozzle without and
with vertical, different height barriers is tested by Schlieren method and the effects are
presented in Fig. 24. The nozzle is designed for an exit Mach number of M=2.6.
2.6.
The classical Schlieren photos obtained with color Schlieren system are presented in
Figs. 25 and 26. The flow around cone with tip angle of 15 and sphere with 40 mm
diameter is tested in a supersonic wind tunnel for different Mach numbers and positions of
color filters.
AE 2751
July-Nov. 2016
28
Flow Visualization
Fig. 22 Black and white Schlieren images in a wind tunnel at (a) M=0.86 and (b) M=1.1
and (c) instantaneous image of bullet and muzzle blast from a 22-caliber rifle
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 23 Colour Schlieren images around (a) a blunt body, (b) a blunt body with a thin protruding
probe and (c) a blunt body with a thin protruding probe at an angle of attack
AE 2751
July-Nov. 2016
29
Flow Visualization
Fig. 24 Visualization of flow in a supersonic nozzle by Schlieren system (flow is

from left to right): Continued_
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 24 Visualization of flow in a supersonic nozzle by Schlieren system (flow is

from left to right): Concluded_
AE 2751
July-Nov. 2016
30
Flow Visualization
(a)
(b)
Fig. 25 Colour Schlieren images obtained in T-36 wind tunnel for around a cone
with 15 degree tip angle at (a) M=1.02 and (b) M =1.56
AE 2751
July-Nov. 2016
Flow Visualization
(a)
(b)
Fig. 26 Colour Schlieren images obtained in T-36 wind tunnel for around a
100 mm dia. sphere at (a) M=1.02 and (b) M =1.56
AE 2751
July-Nov. 2016
31
Flow Visualization
The combined holographic interferometer and Schlieren devices, have
been designed, made and tested for small supersonic and large trisonic wind tunnels.
The device can be included in tests either as Schlieren system or interferometer.
Improvements to the basic Schlieren system include the Rainbow Schlieren,
Schlieren where a colored
bulls eye filter is used rather than a knife edge to quantify the strength of the refraction.
The other variety of Schlieren methods is obtained by using laser as a light source.
Figure 27 illustrates results from a Schlieren system with HeNe laser as a light source.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 27 Schlieren system with laser as light source in T-36 and Schlieren image
around a cone at M=1.1
AE 2751
July-Nov. 2016
32
Flow Visualization
Optical Methods of Flow Visualization: Interferometry
Interferometry is based on the fact that a change in density not only results in a
refraction of the light, but also in a phase shift. In an interferometer parallel light is split
into two beams. One of the beams enters the flow field, the other beam does not enter
the flow field. When both beams are merged and projected on the same photographic
plate, interference occurs when the phase of one of the beams is shifted by a change of
density in the fluid flow.
The most used classical interferometer in the wind tunnel tests is Mach-Zehnder
interferometer (MZI). MZI can be applied to any case of gas flow investigations, where
density difference becomes noticeable as: thermodynamic data, thermal conductivity of
gases, dissociation, aerodynamic application, turbulence, wave or sonic booms.
AE 2751
July-Nov. 2016
Flow Visualization
Optical Methods of Flow Visualization: Holographic Interferometry
Holographic interferometry is an optical method that makes possible complete flow
field testing. The method is non-contact and it does not disturb flow field. It is used for
testing of different objects and phenomena. The greatest advantage of holographic
interferometry, in relation to Schlieren method, is the fact that it provides complete
information in a single plate, allowing a postponement selection of specific types of
flow visualization.
The base of this method is holography, developed in last forty five years. If, on
the some plate, the image of one object is recorded two times in different moments, in
the process of reconstruction both images appear simultaneously and on the same
place in the space. Object waves interfere because of mutually coherence (they
originate
i i t from
f
th same light
the
li ht beam
b
th t illuminate
that
ill i t the
th hologram)
h l
) and
d the
th interference
i t f
effects can be observed in the reconstructed object image. If no change occurs on
object between first and second exposition, then there is no difference in images and
there are no interference fringes. If certain difference appears, then the reconstructed
image contains the system of interference fringes N that indicate that change.
AE 2751
July-Nov. 2016
33
Flow Visualization
Quantitative flow testing, using holographic interferograms is performed by determining
the number of fringes N(x,y) in the field image with respect to a reference point of
known density. After that, the index of light refraction n(x,y) and the air density (x,y)
can be calculated.
calculated For the isentropic flow,
flow relations between N,
N n,
n ,
pressure P,
P
temperature T, velocity V, and Mach number M exist, which van be used to determine
required information.
The simplest case for analysis is the 2D flow. For the processing of interferograms of
axi-symmetrical phase objects, the method of inversion, based on the Abel
transformation, is used. The experiment geometry is usually selected in order to simplify
the mathematical representation of flow and changes occurring at the path of the
laser light beam through the test section.
Computer tomography is an important technique for reconstructing 3-D fields from
holographic interferograms. Several techniques have been developed for computer
tomography as: implicit methods (series expansion, discrete element representations),
explicit methods (convolution method), and Fourier transform method. The choice of the
best algorithm depends on structure of the density field, the amount and format of
available data.
AE 2751
July-Nov. 2016
Flow Visualization
Holographic interferometer with parallel beams is at the same time Schlieren and
shadow device. Figure 28 shows the schematic diagram of the experimental setup.
During the experiments synchronized measurements were performed. Double
exposition technique was used for holographic interferograms recording: wind off
(when homogeneous flow field exists) and wind on (when there is complex flow
field for testing).
Stagnation pressure (P0), atmospheric pressure (Pa), and Mach number (M)
were measured by the primary measurement system (PMS) in the wind tunnel, at
the moment of recording hologram, shadow and/or Schlieren results.
AE 2751
July-Nov. 2016
34
Flow Visualization
Fig. 28 Schematic of holographic interferometry/ Schlieren system and

shadowgraphy in a supersonic wind tunnel
AE 2751
July-Nov. 2016
Flow Visualization
In order to demonstrate advantages of holographic interferometry in complex flow
field testing, and compared with other classical methods, the series of experiments
were performed in MTI supersonic and trisonic wind tunnel at flow velocity from M=0.7
to 3.24. The photos of holographic interferograms illustrate this method. Figure 29
shows some interferograms of different flows.
The use of classical methods of nozzle flow field testing comprises the introduction of
probe within the expansion region and holes perforation on nozzle surface. These
interventions would significantly change the flow field. Optimization of this measurement
is made by the holographic interferometry.
In order to demonstrate and to compare complementary possibilities of optical methods
in quantitative flow visualization, Prandlt-Mayer expansion tested by three optical
methods is presented. Figures 30 a and b show the flow visualization around 90
corner end edge for supersonic nozzle M=1.56. The interferogram is recorded by
double passing, collimated, object beam trough the wind tunnel test section. The
shadowgraph is recoded on a holographic plate, because of collimated beams. The
color Schlieren is recorded in the some time with holographic interferogram.
AE 2751
July-Nov. 2016
35
Flow Visualization
Fig. 29 Holographic interferogram of flow around a (a) missile for M=1.56,

(b) 90 cone at M=0.86 and (c) 2D cylinder at M=0.76 (flow is left to right)
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 30 Visualization of supersonic flow (left to right) around 2D 90O nozzle edge
(Prandtl-Mayer expansion) M=1.56 a) shadowgraph and b) interferogram
AE 2751
July-Nov. 2016
36
Flow Visualization
The holographic interferograms were used for numerical calculation of flow field
parameters in the vicinity of nozzle edge where the expansion fan is formed (Fig. 30c).
The fringe number N was read from this hologram. Points in front of expansion fan have
N=0, since the last fringe has N=17. The theoretical and experimental values of Mach
number in the expansion area are in good agreement Mexp=2.15, Mthe=2.13. The
photos in Figs. 31 a and b present holographic interferograms of flow around a sphere
for M=0.8 (without shock wave) and M=1,06 (bow shock wave is in front of model).
Figure 31 b is combination of holographic interferograms (upper part) and

Schlieren photo for the same flow. On the interferometric part of photo easily seen are:
the stagnation point, the detached bow wave, the vortex sheet generated past sphere
and so on.
AE 2751
July-Nov. 2016
Flow Visualization
(a)
(b)
Fig. 31 Holographic interferogram of flow around a sphere for (a) M=0.82

and (b) mixed hologram and Schlieren for M=1.06
AE 2751
July-Nov. 2016
37
Flow Visualization
Very interesting is example of flow visualization around tunnel wall perforations. Many
transonic tunnels are operated with performed walls in the test section.
A number of investigations have been performed to determine how the flow in the test
section is affected by the presence of the perforation.
perforation The next photos (Fig.
(Fig 32) reports
on test performed in T-36, with a single slanted slot in the bottom plate of the test
section. The disturbances originating from the slot are expressed by distortions of the
parallel fringe system. A concentration of fringes indicated the formation of a pressure
wave. The slanted slot was used because it had been reported that such geometry
would considerably reduce the perturbation of free flow.
The interferogram however shows that the disturbance from the slot is not at all
negligible and reaches even beyond the axis of the test section (to about 60 % of the
test section height). The perturbation has the influence on the model sting in the central
line of the test section (Fig. 32).
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 32 (a) Test section, (b) holographic interferograms of flow (flow is left to right) in the empty wind
tunnel test section with wall perforations (slanted slot) and (c) and with cone for M=0.83
AE 2751
July-Nov. 2016
38
Flow Visualization
The interferograms of several supersonic rocket nozzle configurations (Fig. 33a)
without and with different barriers are recorded in order to provide a good insight in the
physical processes (Figs. 33 b and c).
The theoretical value of Mach number in the output plane of the nozzle is estimated
to be M=2.6. Using the data for pressure measurements, it is obtained M=2.46 and by
means of holographic calculations, Mach number is M=2.56. The placing of barriers in
the supersonic flow, leads to the appearance of the stagnation zone, shock and
expansion waves. Visualization of the flow field made in the experiment indicates strong
interaction of the turbulent boundary layer with the oblique shock wave in the divergent
part of the nozzle.
Beside two-exposition method, there are used the real time method, the average or
sandwich methods,, the speckle
p
interferometry,
y, refraction interferometry,
y, differential
interferometry and so on. Optical holography use laser light in visible spectrum, and
interferential effects are recorded on photo or thermosensitive emulsions. Electronic
holography uses CCD cameras. In some specific cases acoustic and microwave
holography, with electron beam, Xrays, or computer holography can be used. With
similar possibilities speckle interferometry, Moir interferometry and stereography are
used nowadays.
AE 2751
July-Nov. 2016
Flow Visualization
Fig. 33 Holographic interferograms for 2D supersonic nozzle without and with three barriers
AE 2751
July-Nov. 2016
39
Flow Visualization
Optical Methods of Flow Visualization: Infrared Thermography
Thermographic systems have been considered to analyze fluid-dynamic phenomena
thirty years ago. Nowadays high resolution and differential infrared thermographic
measurement systems open up new possibilities in its application.
Temperature field that can be measured by a thermographic system on the surface of a
solid body invested by a flow is determined by many combined effects.
effects Very important
effects are: conversion of kinetic energy of the flow into thermal energy, flow temperature
variation in time and space, convection heat transfer phenomena between flow and body,
conduction phenomena inside the body and radiation heath exchange of the body surface
with surroundings. By correspondence between convective heat transfer coefficient and
local turbulence, it is possible to carry out information about the boundary layer. In
addition to the laminar-to-turbulent transition boundary, the infrared camera was able to
detect shock waves and present a time dependent view of the flow field. Figure 34 shows
thermograms of tests have been performed using an high resolution thermographic
system for fluid-dynamics analysis of a known test case,
case a wing profile,
profile in a wind tunnel
under variable and constant temperature condition at different air flow velocities.
A time dependent heat transfer code was developed to predict temperature
distributions on the test subject and any necessary surface treatment. A commercially
available infrared camera can be adapted for airborne use. Readily available infrared
technology has the capability to provide detailed visualization of various flow
phenomena in subsonic to hypersonic flight regimes.
AE 2751
July-Nov. 2016
Flow Visualization
Optical Methods of Flow Visualization: Infrared Thermography
AE 2751
Fig. 34 Black Aluminum airfoil with (a) incidence of 7.5O clockwise

and (b) incidence of -7.5O clockwise
July-Nov. 2016
40

WIND TUNNEL Notes

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

WIND TUNNEL Notes

Uploaded by

Copyright:

Available Formats

Wind Tunnel Techniques

The course should enable the students to:

Wind Tunnel Techniques

Principles of Model Testing

Buckingham theorem, Non-dimensional numbers, Scale effects, Types of similarities

Classification, Special problems of testing in subsonic, transonic, supersonic and

Calibration of Wind Tunnels

Test section speed, Horizontal buoyancy, Flow angularities, Turbulence

Wind Tunnel Measurements

Pressure and velocity measurements, Force measurements, Three and six

Wind Tunnel Techniques

Flow visualization techniques

Wind Tunnel Techniques

Experimental Fluid Dynamics

One or two for small CFD work,

Usually requires moderate number of

Moderate for computation, but large for

Large for preparation, small for data

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Drag depends on FOUR parameters:

Difficult to know how to set up experiments to determine dependencies

Difficult to know how to present results (four graphs?)

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Experimentally derived relation between the dependent and independent variables

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Exponents to be determined to satisfy dimensional homogeneity

Hence the first dimensionless number can be written as

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

w, and volume flow rate, Q,

which yield dimensionless groups, wr/V and Q/VL .

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Wind Tunnel Techniques

Principles of Model Testing

Fv Shear Stress. Area Shear Stress. Area

Wind Tunnel Techniques

Principles of Model Testing