Non-intrusive sensing of air velocity,
humidity, and temperature using TDLAS
TDLAS: Tunable Diode Laser Absorption Spectroscopy
Suhyeon Park
Mechanical Engineering, Virginia Tech
Advisor
Dr. Lin Ma
Aerospace and Ocean Engineering
Mechanical Engineering Adjunct
Virginia Tech
Outline

Introduction:
Concept of TDLAS measurement

Velocity measurement

Temperature measurement and
non-uniform distribution analysis

Water concentration tomographic
inversion
College of Engineering
2
TDLAS for wind energy
TDLAS sensor
Real-time
monitoring
Research instrument
•
•
•
•
Flow info
Active control of wind turbines
Velocity, humidity and temperature - simultaneous measurement
Non-intrusive in-situ real-time monitoring without particle seeding
Calibration-free accurate measurement
Low cost, low maintenance
College of Engineering
3
Fundamental theory of absorption
spectroscopy
Detector
It()
L
Io ()
L aser
𝐼𝑡 (𝜈)
𝐼𝑜
•
Beer - Lambert relation: 𝜏 𝜈 =
•
Measure 𝑘𝜈 and P to infer T and X : 𝑘𝜈 = 𝑓(𝑇, 𝑋, 𝑃)
College of Engineering
= exp[−𝑘𝜈 ⋅ 𝐿]
4
Principle of TDLAS measurement
•
𝑇 – temperature : inferred from the
shape of spectra (the relative
absorption strengthen of a “cold” and a
“hot” line)
• 𝑋 – H2O concentration : inferred from
the magnitude of the spectra after 𝑇)
• 𝑉 – flow velocity : inferred from Doppler
shift between two beams with an angle
T=2379 K
X=0.347
T=1647
X=0.190
Cold
Line
1.379 mm
College of Engineering
Hot
Line
1.333 mm
5
Outline

Introduction:
Concept of TDLAS measurement

Velocity measurement

Temperature measurement and
non-uniform distribution analysis

Water concentration tomographic
inversion
College of Engineering
6
TDLAS Doppler velocimetry setup
50
𝜃
Laser Diode
1343 nm
90
50
Coupler
50
Coupler
50
Laser Diode
1392 nm
Photodiode 1
Coupler
10
Photodiode 2
MZI
Interferometer
Air Flow
Schematic of H2O absorption velocity measurement
• Tunable diode laser scans across optical frequency
• Mach-Zehnder interferometer converts time series data into frequency
spectrum
• Two beams cross at angle 2𝜃 and Doppler shift is measured
College of Engineering
7
TDLAS velocity measurement
demonstrated in a simple duct
y
Laser
Air Flow
x
Experimental setup of TDLAS velocity
measurement
College of Engineering
Averaged velocity in
the y direction
8
Measured Doppler shift by TDLAS
Voigt Fitting Parameters
Voigt 1
College of Engineering
Voigt 2
y0, offset
-1.2264E-03 -1.3770E-03
xc, center
-1.6779E-03 -1.2582E-03
A, area
4.5947E-03
wG, Gaussian width
-3.6193E-02 -3.4506E-02
wL, Lorenzian width
6.5980E-02
4.6043E-03
6.7691E-02
9
TDLAS measured velocity in
agree with hot wire velocity
• TDLAS velocity agrees well with “averaged” hot-wire velocity
• “Averaging” (i.e., effects of flow non-uniformity) will be more
thoroughly investigated in the future
College of Engineering
10
Outline

Introduction:
Concept of TDLAS measurement

Velocity measurement

Temperature measurement and
non-uniform distribution analysis

Water concentration tomographic
inversion
College of Engineering
11
Ambient air temperature
measurement
Cold Laser Whole Range
Hot Laser Whole Range
1.4
0.03
Simulation
Measurement
1.2
TDLAS
cold line
0.6
0.4
0.01
0.005
0
0
7175
7180
7185
Wavenumber(cm-1)
-0.005
7190
TDLAS
hot line
0.015
0.2
-0.2
7435
7440
7445
7450
Wavenumber(cm-1)
TDLAS
cold
line
Voigt
Fitting
0.025
Measurement
Voigt Fit
0.25
Absorbance
0.15
0.1
Measurement
Voigt Fit
0.02
0.2
Absorbance
7455
TDLAS
line
Voigthot
Fitting
0.3
0.01
•
0
0
Acold = 2.279 e-2
-0.2
cm-1
-0.1
0
0.1
Wavenumber (cm-1)
College of Engineering
Temperature sensing
•
0.015
0.005
0.05
-0.05
Laser controller setting:
• Diode temperature
Tcase,cold = 27.2°C
Tcase,hot = 26.2°C
• Drive current control
Vmod = 1.7 Vpp, ramp signal
0.02
0.8
Absorbance
Absorbance
1
Identification of target lines
Simulation
Measurement
0.025
Room condition
T = 23°C
Measured temperature
TM = 25°C
(error = +2 K)
Ahot 1.307 e-3 cm-1
0.2
-0.005
-0.4
-0.2
0
0.2
Wavenumber (cm-1)
0.4
12
Thermocouple measurement of
flame temperature
McKenna Burner and flame temperature measurement
• Thermocouple was
manually traversed
through a stable flat
flame from a
McKenna burner
1400
• McKenna burner
flame temperature
was about 1200 °C.
Temperature (°C)
1200
1000
800
600
400
200
0
0
College of Engineering
50
Position (mm)
100
13
Setup to demonstrate non-uniform
temperature distribution
• TDLAS measurement was
conducted for premixed
burner flame
Detector
Hot section
Beam path
McKenna
Burner
Laser
Non-uniform temperature distribution setup
College of Engineering
• Goal: To see if TDLAS can
resolve the non-uniform
temperature and water
concentration in the laser
beam path
– Hot section: flame
– Cold sections: ambient air
14
Line-of-sight TDLAS measurement
of Flame Temperature
Line Strength - Temperature Relation
0.05
2
• TDLAS measured temperature
was 936 °C
Acold
Ahot
Ratio
Absorbance
Shot/Scold
Signal
ratio
• Discrepancy is due to
temperature non-uniformity in the
beam path
T
0
500
1000
1500
2000
Temperature (K)
T = 1209 K = 936 °C
College of Engineering
2500
0
3000
• Possible solutions include nonuniform analysis or tomographic
inversion.
15
Flame temperature distribution by
TDLAS non-uniform analysis
TC
TDLAS
1800
1600
1400
1200
T (K)
Case 1
TDLAS 1460 K
TC 1489 K
• Non-uniformly distributed
temperature is found by
thermocouple (TC) and
TDLAS measurement
1000
800
600
400
200
0
20
40
60
80
x (mm)
TC
TDLAS
1800
1600
1400
1200
T (K)
Case 2
TDLAS 1587 K
TC 1588 K
1000
800
• TDLAS measured
temperature distribution
agrees with TC
measurement
600
400
200
0
20
40
60
80
x (mm)
College of Engineering
16
Non-uniform fitting procedure
Temperature Distribution
• Fitting procedure
demonstrated to find T1, X1,
T2, X2 in hot section and cold
sections
Temperature (K)
T1
6 cm
T2
T2
• Valuable for practical
implementation
0
0
– Practical flows are non-uniform
– Undesirable/difficult to mount
sensors near hot flows
Position (cm)
Side View
Flame
Collimator
College of Engineering
McKenna
Burner
Detector
17
Example results from the fitting
procedure
TDLAS absorption spectrum T1 = 1587 K, X1 = 0.117
0.1
0.14
Measurement
Non-uniform
Uniform
0.12
0.08
0.07
Absorbance
0.1
Absorbance
Measurement
Non-uniform
Uniform
0.09
0.08
0.06
0.06
0.05
0.04
0.03
0.04
0.02
0.02
0
0.01
7185.4
7185.5
7185.6
Frequency, cm-1
7185.7
7185.8
0
7444.1
7444.2
7444.3
7444.4
Frequency, cm-1
7444.5
• Our method exploits the shape (i.e., all the data points on the
absorption lines) to obtain distribution information in nonuniform flows
College of Engineering
18
Outline

Introduction:
Concept of TDLAS measurement

Velocity measurement

Temperature measurement and
non-uniform distribution analysis

Water concentration tomographic
inversion
College of Engineering
19
TDLAS setup in high speed jet
measurement
Traverser
Jet
Sensors
Lasers
Laser controller
TDLAS setup installed at the high speed jet facility
Detecter
+ IS
• Laser beams cross at the
jet for TDLAS velocity
measurement with Doppler
effect
Collimator
Beam 1
• Sensors are installed on a
vertical traverser
Beam 2
Detecter
+ IS
• TDLAS measurement
system is installed for high
speed jet measurement
Jet flow
Collimator
TDLAS sensors located at the jet
College of Engineering
20
Water density of high speed jet
obtained by traversing TDLAS
-3
3
x 10
• Line-of-sight averaged
water concentration
measured by traversing
the TDLAS sensor
Jet flow
2.8
XL (A.U.)
2.6
2.4
• Conditions
2.2
Ambient air
Ambient air
2
1.8
-6
-4
-2
0
2
Vertical Position(cm)
4
6
– Ambient air:
25 °C, X = 0.010
– Compressor supplied air:
17 °C, X = 0.008
– Traversing step : 0.4 cm
TDLAS measurement in the high speed jet at Mach 0.65
College of Engineering
21
Projection of non-uniform
distribution
X1
Distribution
5 cm
X1
X2
0
r
y
X1
Front View
X2
• Water density projection
should be reconstructed to
distribution by tomography
X1
Jet flow
Collimator
• TDLAS measurement by
traversing is a projection of
non-uniform distribution
Detector
Projection
0
College of Engineering
y
22
Introduction to Abel inversion
(Tomography in 1D)
l1,1
r
l1,2
l2,2
l3,3
l2,3
l1,3
...
X3
X = A-1P
X2
X1
• Abel inversion is used in
axially symmetric geometry
• The jet is divided into layers
with different water density
(X1, X2, X3 ...)
• Distribution can be obtained
from TDLAS measured
projection by Abel inversion
X - Distribution
A - Geometry
P - Projection
College of Engineering
23
1D tomography applied to
TDLAS jet measurements
-3
16
x 10
Reconstructed X (A.U)
14
• Abel inversion is applied to
TDLAS jet measurements
Jet flow
12
10
8
• Spatial distribution of water
density in the jet is obtained
6
4
2
Ambient air
Ambient air
0
-2
-6
-4
-2
0
r (cm)
2
4
6
Spatial distribution of water density
in the high speed jet at Mach 0.65
College of Engineering
24
Conclusion
• TDLAS measurement of temperature, velocity, water
concentration was demonstrated simultaneously
• Non-intrusive instantaneous measurement capabilities of
TDLAS sensor can be utilized in research facilities and practical
monitoring system of wind energy
College of Engineering
25
Thanks !
Suhyeon Park
Department of Mechanical Engineering
Virginia Tech, Blacksburg, 24060, VA
Tel: (540)750-1559
Email: spark11@vt.edu
College of Engineering
26