Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
HHui Huui HuDepartment of Aerospace Engineering, Iowa State University Department of Aerospace Engineering, Iowa State University
Ames, Iowa 50011, U.S.AAmes, Iowa 50011, U.S.A
Lecture # 13: Lecture # 13: Particle Image Particle Image VelocimetryVelocimetry TechniqueTechnique
AerEAerE 311L & AerE343L Lecture Notes311L & AerE343L Lecture Notes
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
ParticleParticle--based Flow Diagnostic Techniquesbased Flow Diagnostic Techniques
• Seeded the flow with small particles (~ µm in size)
• Assumption: the particle tracers move with the same velocity as local flow velocity!
Flow velocityFlow velocityVVff
Particle velocityParticle velocityVVpp
=
Measurement of particle velocity
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
ParticleParticle--based techniques: Particle Image based techniques: Particle Image VelocimetryVelocimetry (PIV)(PIV)
• To seed fluid flows with small tracer particles (~µm), and assume the tracer particles moving with the same velocity as the low fluid flows.
• To measure the displacements (ΔL) of the tracer particles between known time interval (Δt). The local velocity of fluid flow is calculated by U= Δ L/Δt .
A. t=tA. t=t00 B. t=tB. t=t00+10 +10 μμss C. Derived Velocity fieldC. Derived Velocity fieldX (mm)
Y(m
m)
-50 0 50 100 150
-60
-40
-20
0
20
40
60
80
100
-0.9 -0.7 -0.5 -0.3 -0.1 0.1 0.3 0.5 0.7 0.95.0 m/sspanwise
vorticity (1/s)
shadow region
GA(W)-1 airfoil
t=tt=t00 tLUΔΔ
=
t= tt= t00++ΔΔttΔΔLL
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
PIV System SetupPIV System Setup
Illumination system(Laser and optics)
cameraSynchronizer
seed flow withtracer particles
Host computer
Particle tracers: Particle tracers: to track the fluid movement. to track the fluid movement. Illumination system: Illumination system: to illuminate the flow field in the interest region.to illuminate the flow field in the interest region.Camera: Camera: to capture the images of the particle tracers.to capture the images of the particle tracers.Synchronizer: Synchronizer: the control the timing of the laser illumination and the control the timing of the laser illumination and
camera acquisition.camera acquisition.Host computer: Host computer: to store the particle images and conduct image processing. to store the particle images and conduct image processing.
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Tracer Particles for PIV Tracer Particles for PIV
•• Tracer particles should be Tracer particles should be neutrally buoyant neutrally buoyant andand small enough small enough to follow the flow to follow the flow perfectlyperfectly..
•• Tracer particles should be Tracer particles should be big enoughbig enough to to scatterscatter the illumination lights the illumination lights efficientlyefficiently ..
•• The scattering efficiency of trace particles also strongly depeThe scattering efficiency of trace particles also strongly depends on the ratio of the nds on the ratio of the refractive refractive indexindex of the particles to that of the fluid. of the particles to that of the fluid.
For example: the refractive index of For example: the refractive index of waterwater is considerably larger than that of is considerably larger than that of airair. . The scattering of particles in air is at least one order The scattering of particles in air is at least one order of magnitude more of magnitude more efficient than particles of the same size in water.efficient than particles of the same size in water.
hν
Incident lightScattering light
. d=1μm b. d=10μm c. d=30μm
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
•• A primary source of measurement error is the influence of graviA primary source of measurement error is the influence of gravitational forces when the density tational forces when the density of the tracer particles is different to the density of work fluiof the tracer particles is different to the density of work fluid.d.
•• The velocity lag of a particle in a continuously acceleration flThe velocity lag of a particle in a continuously acceleration fluid will be:uid will be:
Tracer Particles for PIVTracer Particles for PIV
μρ
τ
τ
μρρ
18
);exp(1()(
18)(
2
2
pps
sp
ppPs
d
tUtU
gdUUU
=
−−=
−=−=
rr
rrrUr
gdU ppg μ
ρρ18
)(2 −=
PUr
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
• Tracers for PIV measurements in liquids (water):
• Polymer particles (d=10~100 μm, density = 1.03 ~ 1.05 kg/cm3)
• Silver-covered hollow glass beams (d =1 ~10 μm, density = 1.03 ~ 1.05 kg/cm3)
• Fluorescent particle for micro flow (d=200~1000 nm, density = 1.03 ~ 1.05 kg/cm3).
•Quantum dots (d= 2 ~ 10 nm)
• Tracers for PIV measurements in gaseous flows:
• Smoke …
• Droplets, mist, vapor…
• Condensations ….
• Hollow silica particles (0.5 ~ 2 μm in diameter and 0.2 g/cm3 in density for PIV measurements in combustion applications.
•Nanoparticles of combustion products
Tracer Particles for PIVTracer Particles for PIV
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
illumination systemillumination system
•• The illumination system of PIV is always composed of light sourcThe illumination system of PIV is always composed of light source and optics. e and optics.
•• LasersLasers: such as Argon: such as Argon--ion laser and ion laser and Nd:YAGNd:YAG Laser, are widely used as light Laser, are widely used as light source in PIV systems due to their ability to source in PIV systems due to their ability to emit monochromatic lightemit monochromatic light with with high high energy densityenergy density which can easily be bundled into thin light sheet for illuminatwhich can easily be bundled into thin light sheet for illuminating ing and recording the tracer particles without chromatic aberrationsand recording the tracer particles without chromatic aberrations..
•• OpticsOptics: always consisted by a set of cylindrical lenses and mirrors t: always consisted by a set of cylindrical lenses and mirrors to shape the o shape the light source beam into a planar sheet to illuminate the flow fielight source beam into a planar sheet to illuminate the flow field.ld.
laser optics
Laser beamLaser sheet
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
DoubleDouble--pulsed pulsed Nd:YagNd:Yag Laser for PIVLaser for PIV
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Optics for PIVOptics for PIV
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
CamerasCameras
•• The widely used cameras for PIV: The widely used cameras for PIV:
•• Photographic filmPhotographic film--based cameras based cameras or or ChargedCharged--Coupled Device (CCD) camerasCoupled Device (CCD) cameras..
••Advantages of CCD cameras: Advantages of CCD cameras:
•• It is fully digitizedIt is fully digitized
•• Various digital techniques can be implemented for PIV image proVarious digital techniques can be implemented for PIV image processing.cessing.
•• Conventional autoConventional auto-- or crossor cross-- correlation techniques combined with special correlation techniques combined with special framing techniques can be used to measure higher velocities.framing techniques can be used to measure higher velocities.
•• Disadvantages of CCD cameras:Disadvantages of CCD cameras:
•• Low temporal resolution (defined by the video framing rate): Low temporal resolution (defined by the video framing rate):
•• Low spatial resolution:Low spatial resolution:
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Interlaced CamerasInterlaced Cameras
• The fastest response time of human being for images is about ~ 15Hz• Video format:
– PAL (Phase Alternating Line ) format with frame rate of f=25Hz (sometimes in 50Hz). Used by U.K., Germany, Spain, Portugal, Italy, China, India, most of Africa, and the Middle East
– NTSC format: established by National Television Standards Committee (NTSC) with frame rate of f=30Hz. Used by U.S., Canada, Mexico, some parts of Central and South America, Japan, Taiwan, and Korea.
Even fieldEven field(2,4,6(2,4,6……640)640)
Old fieldOld field(1,3,5(1,3,5……639)639)
Even field
Odd field
16.6ms16.6ms
1 frameF=30Hz
480 pixels by 640 pixels480 pixels by 640 pixels
Interlaced cameraInterlaced camera
timetime
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Progressive scan cameraProgressive scan camera
•• All image systems produce a clear image of the background All image systems produce a clear image of the background •• Jagged edges from motion with interlaced scan Jagged edges from motion with interlaced scan •• Motion blur caused by the lack of resolution in the 2CIF sample Motion blur caused by the lack of resolution in the 2CIF sample •• Only progressive scan makes it possible to identify the driverOnly progressive scan makes it possible to identify the driver
1st frameExp=33.33ms
2nd frameexp=33.33ms
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Synchronizer Synchronizer •• Function of Synchronizer: Function of Synchronizer:
•• To control the timing of the laser illumination and camera acquTo control the timing of the laser illumination and camera acquisitionisition
1stpulsed
2ndpulsed
Timing of pulsed laser
Timing of CCD camera
time
1st frame exposure
2nd frame exposure
Δt33.33ms(30Hz)
To laser To camera
From computer
Synchronizer
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Host computerHost computer
•• To send timing control parameter to synchronizer. To send timing control parameter to synchronizer.
•• To store the particle images and conduct image processing.To store the particle images and conduct image processing.
Host computerTo synchronizer
Image data from camera
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
SingleSingle--frame techniqueframe technique
particleparticleStreak lineStreak line
VVL=V*L=V*ΔΔtt
singlesingle--pulsepulse MultipleMultiple--pulsepulse
Particle streak Particle streak velocimetryvelocimetry
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
MultiMulti--frame techniqueframe techniquea. T=ta. T=t00
b. T=tb. T=t11
c. T=tc. T=t22
a. T=ta. T=t33
t=tt=t00
tLUΔΔ
=
t= tt= t00++ΔΔttΔΔLL
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Image Processing for PIVImage Processing for PIV
•• To extract velocity information from particle images.To extract velocity information from particle images.
t=t0 t=t0+4ms
A typical PIV raw image pair
Y/D
X/D
-2 -1 0 1 2 31
1.5
2
2.5
3
3.5
4
4.5
5
1.1001.0501.0000.9500.9000.8500.8000.7500.7000.6500.6000.5500.5000.4500.4000.3500.3000.2500.2000.1500.100
Velocity U/Uin
Image processing
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Particle Tracking Particle Tracking VelocimetryVelocimetry (PTV)(PTV)
t=t0 t=t0+Δt
Low particle-imagedensity case
1.1. Find position of the particles at each Find position of the particles at each imagesimages
2.2. Find corresponding particle image pair Find corresponding particle image pair in the different image frame in the different image frame
3.3. Find the displacements between the Find the displacements between the particle pairs.particle pairs.
4.4. Velocity of particle equates the Velocity of particle equates the displacement divided by the time interval displacement divided by the time interval between the frames.between the frames.
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Particle Tracking Particle Tracking VelocimetryVelocimetry (PTV)(PTV)--22
Particle position of time step t=t1
Search region for time step t=t4
Search region for time step t=t3
Search region for time step t=t2
FourFour--frameframe--particleparticletracking algorithm tracking algorithm
1.1. Find position of the particles at each Find position of the particles at each imagesimages
2.2. Find corresponding particle image pair Find corresponding particle image pair in the different image frame in the different image frame
3.3. Find the displacements between the Find the displacements between the particle pairs.particle pairs.
4.4. Velocity of particle equates the Velocity of particle equates the displacement divided by the time interval displacement divided by the time interval between the frames.between the frames.
PTV results
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
CorrelationCorrelation--based PIV methodsbased PIV methods
high particle-image density
t=t0 t=t0+Δt Corresponding flow velocity field
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
CorrelationCorrelation--based PIV methodsbased PIV methods
t=t0 t=t0+Δt
Searching window size (SB)
SB
SA
SA Interrogation window
SB
q(x,y) P(x,y) SB
( )dvgyxgdvfyxf
dvgyxgfyxfqpR
22)),(()),((
)),(()),((,
∫∫∫
−−
−−=Correlation coefficient
function
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Cross Correlation OperationCross Correlation Operation
Signal A: Signal A:
Signal B: Signal B:
( )dxuxgdxxf
dxuxgxfuR
])([*])([
)](*)([22
∫∫∫
+
+=
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
1 4 7 10 13 16 19 22 25 28 31S1
S10
S19
S280.7
0.75
0.8
0.85
0.9
0.95
1
Correlation coefficient distributionCorrelation coefficient distribution
( )dvgyxgdvfyxf
dvgyxgfyxfqpR22
)),(()),(()),(()),((,
∫∫
∫
−−
−−=
R(p,q)Peak location
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Comparison between PIV and PTVComparison between PIV and PTV•• Particle Tracking Particle Tracking VelocimetryVelocimetry::
•• Tracking individual particleTracking individual particle•• Limited to low particle image density caseLimited to low particle image density case•• Velocity vector at random points where tracer particles exist.Velocity vector at random points where tracer particles exist.•• Spatial resolution of PTV results is usually limited by the Spatial resolution of PTV results is usually limited by the
number of the tracer particlesnumber of the tracer particles
•• CorrelationCorrelation--based PIV:based PIV:•• Tracking a group of particlesTracking a group of particles•• Applicable to high particle image density caseApplicable to high particle image density case•• Spatial resolution of PIV results is usually limited by the sizeSpatial resolution of PIV results is usually limited by the size
of the interrogation window sizeof the interrogation window size•• Velocity vector can be at regular grid points.Velocity vector can be at regular grid points.
PTV
t=t0+Δt
PIV
Y/D
X/D
-2 -1 0 1 2 31
1.5
2
2.5
3
3.5
4
4.5
5
1.1001.0501.0000.9500.9000.8500.8000.7500.7000.6500.6000.5500.5000.4500.4000.3500.3000.2500.2000.1500.100
Velocity U/Uin
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Estimation of differential quantitiesEstimation of differential quantities
X (mm)Y
(mm
)10 15 20 25 30 35
5
10
15
20
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Estimation of differential quantitiesEstimation of differential quantities
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Estimation of Estimation of VorticityVorticity distributiondistribution
yU
xV
z ∂∂
−∂∂
=ϖ
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Estimation of Estimation of VorticityVorticity distributiondistribution
Stokes Theorem: Stokes Theorem:
dA
AdldV
yxz
C S
−Γ=⇒
•−=−=Γ ∫ ∫∫
ϖ
ϖrvrr
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
VorticityVorticity distribution Examplesdistribution Examples
-50 0 50 100 150 200 250 300-50
0
50
100
150
200-25.00 -20.00 -15.00 -10.00 -5.00 0.00 5.00 10.00 15.00 20.00 25.00
Spanwise Vorticity ( Z-direction )
Re =6,700
Uin = 0.33 m/s
X mm
Ym
m
Uou
t
water free surface
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 -3.2 -2.7 -2.2 -1.7 -1.2 -0.7 -0.2 0.3 0.8 1.3 1.810 m/sspanwise
vorticity (1/s)
shadow region
GA(W)-1 airfoil
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
EnsembleEnsemble--averaged quantitiesaveraged quantities
•• Mean velocity components in x, y directions: Mean velocity components in x, y directions:
••Turbulent velocity fluctuations:Turbulent velocity fluctuations:
•• Turbulent Kinetic energy distribution:Turbulent Kinetic energy distribution:
•• Reynolds stress distribution:Reynolds stress distribution:
NUuuN
ii /)('
1
2∑=
−= ∑=
−=N
ii Vvv
1
2)('
∑=
=N
ii NuU
1/ ∑
=
=N
ii NvV
1/
)''(21 22
vuTKE += ρ
∑=
−−−=−=
N
i
ii
NUvUuvu
1
))(('' ρρτ
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
EnsembleEnsemble--averaged quantitiesaveraged quantities
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 U m/s: -1.0 1.0 3.0 5.0 7.0 9.0 11.0 13.0 15.0
10 m/s
shadow region
GA(W)-1 airfoil
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 vort: -3.0 -2.0 -1.0 0.0 1.0 2.0 3.0
10 m/s
shadow region
GA(W)-1 airfoil
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 -0.035 -0.025 -0.015 -0.005 0.005 0.015 0.025 0.035
shadow region
GA(W)-1 airfoil
NormalizedReynolds Stress
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 0.010 0.020 0.030 0.040 0.050 0.060 0.070 0.080 0.090 0.100
shadow region
GA(W)-1 airfoil
T.K.E
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Pressure field estimationPressure field estimation
)(1
)(1
2
2
2
2
2
2
2
2
yv
xv
yp
yvv
xvu
yu
xu
xp
yuv
xuu
∂∂
+∂∂
+∂∂
−=∂∂
+∂∂
∂∂
+∂∂
+∂∂
−=∂∂
+∂∂
ρμ
ρ
ρμ
ρ
Copyright Copyright ©© by Dr. Hui Hu @ Iowa State University. All Rights Reserved!by Dr. Hui Hu @ Iowa State University. All Rights Reserved!
Integral Force estimationIntegral Force estimation
FVdfAdPAdVVVdVt VCSCSCVC
rrrrrrr++•=•+
∂∂
∫∫∫∫........
~)( ρρρ
X (mm)
Y(m
m)
-20 0 20 40 60 80 100 120 140
-60
-40
-20
0
20
40
60 U m/s: -1.0 1.0 3.0 5.0 7.0 9.0 11.0 13.0 15.0
10 m/s
shadow region
GA(W)-1 airfoil