Slides_particle Image Velocimetry

  • Upload
    kaffel

  • View
    222

  • Download
    0

Embed Size (px)

Citation preview

  • 8/18/2019 Slides_particle Image Velocimetry

    1/18

    1

    Measurement Techniques

    Digital Particle Image Velocimetry

    Heat and Mass Transfer Laboratory (LTCM)

    Sepideh Khodaparast

    Marco Milan

    Navid Borhani

    Spring semester 2011

  • 8/18/2019 Slides_particle Image Velocimetry

    2/18

    2

    !  Introduction

    !  Particle Image Velocimetry features

    System components for Particle Image Velocimetry

    Principles of Particle Image Velocimetry

    Practical test (TPA) description

    Content

    Spring semester 2011

  • 8/18/2019 Slides_particle Image Velocimetry

    3/18

    3Lausanne, 08/03/2010

    Origins of Particle Image Velocimetry

    Ludwig Prandtl water tunnel (1904)  Poohsticks (1928) 

  • 8/18/2019 Slides_particle Image Velocimetry

    4/18

    4

    Particle Image Velocimetry applications

    Spring semester 2011

    Fluid velocity measurement for fluid dynamic

    characterization:  Air flowing around a car or anaircraft, water running through hydroelectric turbine,

    blood flow, etc.

  • 8/18/2019 Slides_particle Image Velocimetry

    5/18

    5

    Particle Image Velocimetry features

    Spring semester 2011

    Advantages:

    Non-intrusive technique: no modification of the flowproperty at the scale of interest 

    "  Good resolution and accuracy: instantaneous velocityvector maps in a cross-section of the flow

    2D-3D Velocity field reconstruction: 3D componentsmay be obtained with the use of a stereoscopicarrangement

    Drawbacks:

    Setup Time: need to optimize a number of parameters"  Cost 

  • 8/18/2019 Slides_particle Image Velocimetry

    6/18

    6

    System components for Particle Image Velocimetry

    Spring semester 2011

    Test section, seeding particles

    "  Laser sheet 

    "  Camera

  • 8/18/2019 Slides_particle Image Velocimetry

    7/18

    7

    Principles of Particle Image Velocimetry

    Spring semester 2011

    FlowDirection

    SeedingparticlesTube Wall

    Time: tPosition: (x1, y1)

    Time: t + !tPosition: (x2, y2)

    Velocity vector at:

    !"#$

    %&

    '(

    '(=!

    "#$

    %&   ++

    '+

      t 

     y y

     x x y y x x

    t t 

    1212

    2

    2121 ,2

    ,2

  • 8/18/2019 Slides_particle Image Velocimetry

    8/18

    8

    Cross-correlation function

    Spring semester 2011

     !    +=   dx s x I  x I  s R   )()()( 21

    "  I1 and I2 are sub-area (interrogation windows) of the total frame

    "  x is the interrogation location

    s is the shift between the images

    ( ) ( ) ( ) y y x x I  y x I  y x R x y

    !+!+=!!   ""   ,,, 21

    Two-dimensional discrete correlation function: 

    R(!"# !"%& '())*+,-(. /,0 

  • 8/18/2019 Slides_particle Image Velocimetry

    9/18

    9

    Particle Image Velocimetry protocol

    Spring semester 2011

    1 - Divide the images into a regular grid of

    smaller regions: interrogation windows (IW)

    2 - Each IW of the first image is correlated with

    the corresponding IW of the second image

    3 - Find the location of the displacement peak

    and compute the velocity vector

    4 - Reconstruct the flow velocity field

  • 8/18/2019 Slides_particle Image Velocimetry

    10/18

    10

    Particle selection

    Spring semester 2011

    Capturing fluid motion

    "  Small enough to follow fluid motion

    " Large enough to be visible

    "  Homogeneously distributed

    "  Tracer should not alter fluid / flow properties

    Stokes number: F 

    v

     flowof   sticcharacteritime

    timeresponse particleSt 

    !  

    !  

    ==

     _  _  _ 

     _  _ 

    Stv1: The particles will be unaffected by the fluid 

    Optimal particle image diameter: DI =2.5 pixels 

  • 8/18/2019 Slides_particle Image Velocimetry

    11/18

    11Spring semester 2011

    Seeding particle density(number of particle per interrogation window)

    2

    2

    0

    0

     I  I   D M 

     z C  N 

      !=

    C Particle concentration!12 Light sheet ticknessDI  Interrogation window sizeM0  Magnification

    More particles: better signal to noise ratioUnambiguous detection of peak from noise

    NI = 10

    (10 particles per IW are sufficient to perform PIV)

    NI = 5  NI = 10 NI = 15 

  • 8/18/2019 Slides_particle Image Velocimetry

    12/18

    12Spring semester 2011

    Maximum in-plane particle displacement

    !345I = 0.00FI=1.00 

    0.28 0.56 0.850.64 0.36 0.16 

    !3 Particle displacement

    DI  Interrogation window sizeFI In plane loss-of-correlation

    X,Y Displacements < quarter of the interrogation window size

    Choose the sampling interval and the optical magnification factor so thatthe maximum image displacement is less than a quarter of the

    interrogation window size 

  • 8/18/2019 Slides_particle Image Velocimetry

    13/18

    13Spring semester 2011

    Maximum out-of-plane particle displacement

    !64 !10 = 0.00Fo=1.00 

    0.25 0.50 0.750.75 0.50 0.25 

    !6 Particle displacement normal to the observed plane

    !12  Light sheet thicknessFo Out-of-plane loss-of-correlation

    Z Displacement < quarter of the light sheet thickness

    Define a suitable observation plane and choose a sampling interval sothat the particle shift normal to the observed plane is less than a

    quarter of the light sheet thickness 

  • 8/18/2019 Slides_particle Image Velocimetry

    14/18

    14Spring semester 2011

    Spatial gradient - Interrogation window size

    The PIV resolution is proportional to

    the IW size. It has to be defined to

    resolve local flow gradients

    Keeping constant the particle size and density, altering the IW size will

    change the number of particles used to calculate the local velocity in an

    interrogation window. The total number of vector for a given image pair

    depends directly on the IW size

  • 8/18/2019 Slides_particle Image Velocimetry

    15/18

    15Spring semester 2011

    Interrogation window overlap

    Overlap of the interrogation windows increase the number of particle

    that are used in calculating the flow field

    "  Better spatial resolution

    - Longer computational time

    IW 1 IW 2

    50% IW Overlap

    Interrogation window offset

    Offset increase the PIV accuracy for flow with a dominant direction. The

    IW is shifted between the first and the second image

    Better accuracy

    - Longer computational timeIW Offset

    Image 1  Image 2 

    IW 1 IW 1 

  • 8/18/2019 Slides_particle Image Velocimetry

    16/18

    16

    PIV Design rules 

    Spring semester 2011

    Optimal particle image diameter

    Image density

    In-plane motion

    Out-of-plane motion

    DI =2.5 pixels

    NI = 10

    !!X ! < 0.25 IW Size

    !!z ! < 0.25 light sheet thickness

    The IW size and the sampling interval define the spatial and thetemporal resolution 

  • 8/18/2019 Slides_particle Image Velocimetry

    17/18

    17Spring semester 2011

    Practical test (TPA) description

    1: Laminar flow

    2: guidingvanes

    3: elbow

    PIV system components:

    "  Test section

    Laser sheet"  Camera 

    TP procedure:

    1.  Tune and calibrate the PIV system

    2. 

    Record images at 3 differentlocations

    3.  Process the images on the PC

    4. 

    Save and export the results

    TP objectives:

    PIV system tuning (Record nice images)

    "  Understand the principles of the PIV technique (ex: correlation function)

    and the influence of the measurement parameters (n° of particles, size of

    the IW, overlap, etc.) 

  • 8/18/2019 Slides_particle Image Velocimetry

    18/18

    18Lausanne, 08/03/2010

    Enjoy the TPA !!!