AerE 344 Lecture Notes

Lecture #06 Hotwire anemometry: Fundamentals

and instrumentation

Dr. Hui Hu

Department of Aerospace Engineering

Iowa State University
Ames, Iowa 50011, U.S.A

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Technical Fundamentals -1
• Thermal anemometers:
• Measure the local flow velocity through its relationship to the convective cooling of electrically
heated metallic sensors.

• Hot wire anemometers:

• for clean air or other gas flows

• Hot film anemometers:

• for liquid or some gas flows

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 How a Hot wire Sensor Works
The electric current (i) flowing through the wire
generates heat (i2Rw)

Flow Field
V

In equilibrium, this must

be balanced by heat lost
(primarily convective) to
the surroundings.

Electric current, i, • Price: ~$2750

through wire

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Technical Fundamentals
• Heat transfer characteristics:
• Convection (nature convection, forced convection
or mixed convection depending on Richardson 
Fluid flow
numbers) 
• Conduction to the supporting prong V ,T
• Radiation: <0.1%, is negligible. Hot wire Tw
q Tw  T
Nu =
lk (Tw − T )
= Nu(Re, Pr, Gr, M , Kn, aT , l / d , )

Ud 
Re = ; Pr =
 
g (Tw − T )d 3
V
Gr = ; M =
2 c
 1 M
Kn = = c p / cv
d 2 Re
T −T
aT = w
T prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Technical Fundamentals
1
Following King’s Law (1915), Nu = ( A + B Re n )(1 + aT ) m
2
1
Nu = (0.24 + 0.56 Re 0.45 )(1 + aT ) 0.17 , for 44  Re  140
2
According to Collis and Willams (1959): 1
Nu = 0.48 Re 0.51 )(1 + aT ) 0.17 , for 0.02  Re  44
2

For a given sensor and fixed overheat ratio, The above equation can transfer as the relationship
between the voltage output, E, of the hot-wire operation circuit and the flow velocity

E
= A + BV n
Tw − T
Wire temperature cannot be measured directly, but can be estimated from its relationship to the
wire resistance, Rw, directly measured by the operating bridge.
For metallic wires:

Rw = Rr [1 + ar (Tw − T )]
ar : thermal resistivity coefficient
Tr : reference temepature
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Technical Fundamentals

Flow Field The hot wire is electrically heated.

If velocity changes for a unsteady

flow, convective heat transfer changes,
wire temperature will change and
V eventually reach a new equilibrium.

The rate of which heat is

removed from the sensor is
Current flow directly related to the
through wire velocity of the fluid
flowing over the sensor
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Technical Fundamentals
• For a sensor placed in a unsteady flow, the unsteady energy equation will
become:
dTw
mc = i 2 Rw − q (V , Tw )
dt
m : the mass of the sensor
c : specifich heat of the sensor
q : convective heat flux q = q (V , Tw )
The above equation has three unknowns: i, Tw (or Rw) and V

To render this equation solvable, one must keep with the electric current, i, or the
sensor temperature (Tw) constant, which can be achieved with the use of suitable
electric circuits.

The corresponding methods are known as:

(1). Constant-current anemometry
(2). Constant-temperature anemometry

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Constant-current anemometry
Voltage follower
C
Rs
Rs  Rw E
Ec
i = Eo /( Rs + Rw )
Rc
 Eo / Rs  const
Rw
The voltage output will be Eo sensor Compensation R’c
E = i • Rw circuit

The unsteady energy equation is highly-nonlinear. When linearized in the vicinity of an operation
point, namely at a particular flow speed, Vop, and sensor temperature, Twop , it leads to the following
. first-order differential equation: dT
w w
+ (Tw − Twop ) = K T (V − Vop )
dt
w : a time constant, which is proportional to the overheat ratio, and a static sensitivity, KT

Since voltage, E, is proportional to, Rw , which, in turn, is linearly related to Tw, the linearized E-V
relationship will be: dE
w + ( E − Eop ) = K (V − Vop )
dt
 w : is usually ~ 1ms for thin hot-wire and ~ 10 ms for slim cylindrical hot-film.
For flow with variable velocity or temperature, overheat ratio will vary as well.
Flow low speed flow, it may result in “burnout”, for high-speed flow, sensitivity is low
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Constant-current anemometry
Voltage follower
C
Rs
Rs  Rw E
Ec
i = Eo /( Rs + Rw )
Rc
 Eo / Rs  const
Rw
The voltage output will be Eo sensor Compensation R’c
E = i • Rw circuit

The unsteady energy equation is highly-nonlinear. When linearized in the vicinity of an operation
point, namely at a particular flow speed, Vop, and sensor temperature, Twop , it leads to the following
. first-order differential equation: dT
w w
+ (Tw − Twop ) = K T (V − Vop )
dt
w : a time constant, which is proportional to the overheat ratio, and a static sensitivity, KT

Since voltage, E, is proportional to, Rw , which, in turn, is linearly related to Tw, the linearized E-V
relationship will be: dE
w + ( E − Eop ) = K (V − Vop )
dt
 w : is usually ~ 1ms for thin hot-wire and ~ 10 ms for slim cylindrical hot-film.
For flow with variable velocity or temperature, overheat ratio will vary as well.
Flow low speed flow, it may result in “burnout”, for high-speed flow, sensitivity is low
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Constant-temperature anemometry (CTA) - 1

• Electric current through the sensor is adjustable continuously through an

electric feedback system, and in response to the changes in convective
cooling, to make the temperature of the hot wire keep in constant.
• The unsteady energy equation becomes steady equation
• Dynamic response of the anemometer is the same as its static response with
a wide frequency range.
.

dTw
mc = i 2 Rw − q (V , Tw )  i 2 Rw − q (V ) = 0
dt

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Constant-temperature anemometry (CTA)-2

R1
Esw
RR2
2
Eoffset
Rsw
Ew - E
EB
+
Rd Rw Differential
sensor amplifier

Constant temperature circuit

• Sensor, Rw, comprises one leg of the Wheatstone bridge.
• An adjustable decade resistor array, Rd, compress opposite leg of the bridge.
• The bridge ratio R2/R1 is fixed, and R2/R110~20 to make sure to supply most of the available power
to the sensor.
• The two midpoints of the bridge are connected the input of a high-gain, low noise differential
amplifier, whose out put is fed back to the top of the bridge.
• If R2/Rd= R1/Rw, then EB-Ew=0, the amplifier output will be zero.
• If Rd is increased to a value R’d, the resulting bridge imbalance will generate an input imbalance to
the amplifier.
• The amplifier will create some current through both legs of the bridge. The additional current
through the hot wire will create additional joule heating, which tend to increase its temperature and
thus its resistance, until the resistance increasing sufficiently to balance the bridge once more.
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Velocity orientation effects:

– Effective cooling velocity
Veff = V cos .

Fluid flow
– In reality, flow velocity 
tangential to the sensor would V ,T
result in cooling. Hot wire Tw
– Veff = V (cos2 + k2 sin2 )1/2
Tw  T
– Typical values of K2 are 0.05
and 0.20.

prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Prong interference effects:

– Interference of the prongs and the probe
body may produce additional

Fluid flow
complications of the heat transfer 
characteristics. V ,T
– For example, a stream in binormal
Hot wire Tw
direction will produce higher cooling Tw  T
than a stream with the same velocity
magnitude but in the normal direction.
– In reality,Veff = (VN2+ K2 VT2 + h2 VB2 )1/2
– VN , VT and VB are the normal tangitail
and binormal velocity components.
– Typically, h2=1.1~1.2
– To minimize the effect, it usually use
long and thin prongs. Tapered prongs
are also recommended.
prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Heat conduction effects:

– Previous analysis is based on 2-D
assumption with l/d =  .

Fluid flow
– In reality, the effect of end conduct may

effect the accuracy of the measurement
V ,T
results
Hot wire Tw
– Cold length, lc = 0.5*d ((Kw2/K)(1+aR)/Nu) 1/2 Tw  T
– Kw is thermal conductivity of the sensor
– K is thermal conductivity of the fluid
– aR is overheat ratio
– Effect of the sensor length l/lc
– A recent study has demonstrate that end
conduction effects are expected to
decrease significantly as the Reynolds
number increasing

prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Compressibility effects:
– The velocity and temperature fields around
the sensor become quite complicated

Fluid flow
when M>0.6. 
V  SV
V ,T
  S
Hot wire Tw
T0  ST0 Tw  T
For M  1.2 SV = S 

Modified King’s law for compressible flow:

E 2 = A + B( V ) n
n  0.55

prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Temperature variation effects:

– Calibration at Temperature T1.

– Correlation is needed if real Fluid flow
measurements will be conducted at

Temperature T2.
V ,T
Hot wire Tw
– When the flow temperature varies from
position to position or contain turbulent Tw  T
fluctuations, corrections is much more
complicated.
– It requires simultaneous flow temperature
measurements.
– Sv is increasing with overheat ratio aT.
– At extremely low aT, a thermal
anemometer is totally insensitive to
velocity variations, and becomes a
resistance thermometer. The sensor is
called cold wire. prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Composition effects:
– Composition of flow may affect the
convective heat transfer from a thermal

Fluid flow
anemometer in as much as it affect the 
heat conductivity of surrounding fluid. V ,T
– It requires simultaneous measurements of
Hot wire Tw
fluid species concentration. Tw  T

According to Collis and Willams (1959):

1
Nu = (0.24 + 0.56 Re 0.45 )(1 + aT ) 0.17 , for 44  Re  140
2
1
Nu = 0.48 Re 0.51 )(1 + aT ) 0.17 , for 0.02  Re  44
2

prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Various effects and error source

• Reverse flow and high-turbulence

effects:

– thermal anemometer could not resolve Fluid flow
velocity orientation.

V ,T
– Forward flow can not be identified from Hot wire Tw
reversing flow
– In highly turbulent flow (turbulent
Tw  T
intensity >25%), reverse flow will occurr
statistically some time, therefore, using
thermal anemometer for the flow velocity
measurement may result quite large
measurement uncertainty.
– Pulsed Hot –wire concept

prongs
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Multi-sensor probes
• Cross-wire (X-wire) design:

Veff-A
2
Veff − A = (V1 + V2 )
2 V2
2
Veff − B = (V1 − V2 ) V1
2

V
V2
2 V2
V1 = (Veff − A + Veff − B ) V1
V1
2
2
V2 = (Veff − A − Veff − B ) Veff-B
2

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Multi-sensor probes
• Three sensor design
• Four sensor design:

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Diameter of hot wires

• L = 0.8 ~ 1.5 mm
• D = ~ 5 m for conventional applications
• D = ~ 10 m for high-speed applications
• D = ~ 2 m for low speed applications
• Prongs: usually tapered to be d 1mm

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

Lecture #05 Determination of the Aerodynamic Performance of a
Low Speed Airfoil based on Pressure Distribution Measurements

Dr. Hui Hu
Department of Aerospace Engineering
Iowa State University
Ames, Iowa 50011, U.S.A

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Aerodynamic Performance of An Airfoil
60

40

20

Y /C *100
dCl
= 2
0
Before stall
• Basic Thin Airfoil Theory d
GA(W)-1 airfoil
-20

1.4
25 m/s shadow region
L
Cl = -40
1
V 2 c
1.2

2 vort: -4.5 -3.5 -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 4.5
-60
1.0
Lift Coefficient, Cl

Airfoil stall -40 -20 0 20 40

X/C *100
60 80 100 120 140
0.8

60
0.6
CL=2
Experimental data
0.4 40

0.2
0 2 4 6 8 10 12 14 16 18 20
20
0.40
D Angle of Attack (degrees)
Cd =
Y /C *100

1
0.35
V 2 c 0
2
0.30
GA(W)-1 airfoil After stall
Drag Coefficient, Cd

0.25 Experimental data

-20
25 m/s
0.20 shadow region
0.15 -40

0.10 Airfoil stall vort: -4.5 -3.5 -2.5 -1.5 -0.5 0.5 1.5 2.5 3.5 4.5

0.05 -20 0 20 40 60 80 100 120

X/C *100
0
0 2 4 6 8 10 12 14 16 18 20

Angle of Attack (degrees)Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Determination of the Aerodynamic Performance of a Low Speed
Airfoil based on Pressure Distribution Measurements

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

Determination of the Aerodynamic Performance of a Low Speed
Airfoil based on Pressure Distribution Measurements

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Determination of the Aerodynamic Performance of a Low Speed
Airfoil based on Pressure Distribution Measurements
-3.0

-2.5
upper surface

P − P
-2.0 lower surface

Cp =
-1.5

-1.0

Cp
Before stall
1 -0.5

V  2
0.5
0

2 1.0

1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

x/c

• GA(W)-1 airfoil model with 43 pressure tabs

-1.5

upper surface
-1.0 lower surface

42 26
35 21 -0.5

1 Cp
0
12
0.5
5 17
After stall
1.0

1.5
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0

x/c

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

 Determination of the Aerodynamic Performance of a Low Speed
Airfoil based on Pressure Distribution Measurements
What you will have available to you for this portion of the lab:
• A Pitot probe already mounted to the floor of the wind tunnel
for acquiring dynamic pressure throughout your tests.
• A Setra manometer to be used with the Pitot tube to measure
the incoming flow velocity.
• A thermometer and barometer for observing ambient lab
conditions (for calculating atmospheric density).
• A computer with a data acquisition system capable of
measuring the voltage from your manometer.
• The pressure sensor you calibrated last week
• A NACA 0012 airfoil that can be mounted at any angle of
attack up to 15.0 degrees.
• Two 16-channel Scanivalve DSA electronic pressure scanners.

• GA(W)-1 airfoil model with 43 pressure tabs

42 26
35 21
1
12
5 17

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!

Determination of the Aerodynamic Performance of a Low Speed
Airfoil based on Pressure Distribution Measurements
• Calculating airfoil lift coefficient and drag coefficient by numerically integrating the
surface pressure distribution around the airfoil:

 i +1 / 2 2 ( pi + pi +1 )
1
p =

 p N +1 / 2 = ( p N + p1 )
1
 2
xi = xi +1 − xi , y i = y i +1 − y i
x = x − x , y N = y1 − y N
 N 1 N

A'i = − pi +1/ 2 yi

N 'i = pi +1/ 2xi
N N
N ' =  N 'i =  pi +1/ 2 xi (6)
i =1 i =1
N N
A' =  A'i = − pi +1/ 2 yi (7)
i =1 i =1

L' = N ' cos  − A' sin 

D' = N ' sin  + A' cos 
Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
 Required Plots for the Lab Report

• You must generate plots of CP for the upper and lower surfaces of the
airfoil for the angles of attack that you tested.
• Make comments on the characteristics of the CP distributions.
• Calculate CL and CD by numerical integration CP for the angles of attack
assigned to your group.
• You must report the velocity of the test section and the Reynolds number
(based on airfoil chord length) for your tests.
• You must provide sample calculations for all the steps leading up to your
final answer.
• You should include the first page of the spreadsheet used to make your
calculations

Copyright © by Dr. Hui Hu @ Iowa State University. All Rights Reserved!