Professional Documents
Culture Documents
BR Tomographic PIV
BR Tomographic PIV
Volumetric (3D)
Imaging
References:
[1] Scarano, Meas. Sci. Technol. 24, 2013
[2] Violato et al., Proc. ASME Engineering Conf. Hamamatsu, Japan, 2011
2
Tomographic PIV*
Tomo-PIV
Principle
Tomographic reconstruction and 3D cross-correlation of voxel
intensities
Tracer particles within the measurement volume are illuminated by a high power
pulsed light source and the scattered light pattern is recorded simultaneously
from typically 4 viewing directions using high resolution cameras 1 .
The 3D particle distribution is reconstructed by a tomographic reconstruction
algorithm (Multiplicative Algebraic Reconstruction Technique, MART) as a 3D
light intensity distribution for each voxel 2 .
The particle displacement within a chosen interrogation volume is then obtained by 3D cross-correlation of the reconstructed particle distribution at two
exposures separated by a short time step t using advanced iterative multi-grid
algorithms with deformed interrogation volumes 3 .
pixel line-of-sight
voxels
pixels
CCD 1
CCD 3
CCD 2
intensity field
acquire projections
image volumes
reconstruction
cross-correlation
Volume
Self-Calibration*
High precision camera calibration for
3D reconstruction
A precise volumetric calibration based on a calibration target is required
for all tomographic imaging applications.
For Tomo-PIV the calibration accuracy needs to be better than 0.1 pixel
throughout the measurement volume [1, TU-Delft].
Challenge
Camera mounts or other mechanical parts are never 100% stiff.
Temperature-changes cause extension or contraction of mechanical
parts, wind loads or other vibration sources effect the camera adjustment. Unnoticed, in many experiments, the initial volume calibration
will become inaccurate at the time of measurement.
Cam 1
Cam 3
Cam 2
Solution
LaVisions patented* Volume Self-Calibration [2, LaVision allows the
detection and correction of calibration inaccuracies using the actual tracer particle recordings. Misalignments are corrected and the required
calibration accuracy is recovered.
Volume Self-Calibration
Due to inaccuracies in the calibration function, particles in the volume
are imaged at slightly offset positions in the camera images. Averaging these differences for many particles in a local sub-volume, 3D
disparity maps are generated and the calibration function is corected
accordingly.
This widely accepted Volume Self-Calibration procedure provides a
check for and a remedy of possible calibration problems and is an
References:
[1] Elsinga et al., Exp. Fluids 41, 933-947, 2006
[2] Wieneke, Exp. Fluids 45, 549-556, 2008
Fluid
in Motion
Turbulent fluid motion is the dominating flow configuration for nearly
all natural flows as well as for technical applications.
Details of such complex fluid motion can now be quantitatively visualized
in 3D (4D) using (time-resolved) Tomo-PIV e. g. for the validation of
CFD-codes. Thus Tomo-PIV marks an important step in experimental fluid
mechanics research providing insight into previously unseen fluid dynamic phenomena.
reattachment
Tomo-PIV
2 camera Tomo-PIV
Applying the new reconstruction method Motion Tracking Enhancement
(MTE) for time-resolved recordings a reduced 2 camera setup already
enables Tomo-PIV measurements.
This approach allows the direct migration from Stereo-PIV (2D3C) to full
volumetric Tomo-PIV measurements (3D3C).
DaVis
Easy Reliable - Approved
DaVis - the complete software solution
DaVis accompanies all steps of tomographic PIV measurements in a single integrated software: camera and
laser control, external triggering, recording, calibration,
data analysis, display and data export.
With the built in processing history, you are able to reveal recording
settings and previous processing steps to keep track on your datas
history. The project manager keeps all camera images, the calibration and processing results in a manageable structure.
Tomo PIV
patent filed
2003
First
paper
2006
MART
FFT Correlation
2007
Volume Self
Calibration
2008
Motion Tracking
Enhancement
2009
Fast MART
Fast Direct Correlation
2011
Algorithms
DaVis Tomo-PIV:
an ongoing process of innovations
The first publication in 2006 triggered a real gold rush in the scientific
community.
This rush is best documented by the ever growing number of publications,
dealing with the development and, more and more, with the application
of tomographic PIV. Taking an active part in the ongoing process of innovation, our strategy is to make the benefits of innovation available in the
easy-to-use environment of DaVis.
Major milestones in this innovative process, that directly found their way
in the DaVis software, include:
4significant reduction of computation time using sparse reconstruction and direct correlation
4utilizing the computational power of modern graphic cards (Tomo-PIV on GPU)
4robust Tomo-PIV in situations with obstacles and reflections (advanced volume masking)
4taking advantage of time-resolved data for significant noise reduction and the calculation of acceleration
(Fluid Trajectory Correlation, similar to sliding sum of correlation)
References:
[1] Schanz et al. (2013) Meas. Sci. Technol. 24 024009
[2] Wieneke B (2013) Meas. Sci. Technol. 24 024008
[3] Schanz et al. 10th Int. Symp. on PIV, 2013, Delft, The Netherlands
Tomo-PIV on
GPU
2012
Advanced Volume
Masking
Fluid Trajectory
Correlation
2013
Optical Transfer
Function
2014
Shake The
Box
Biolocomotion
Reference:
[1] Bomphrey et al., J. R. Soc. Interface 9, 3378-3386, 2012
10
Tomo-PIV
2 mm large Daphnia
3D reconstruction of
swimming Daphnia
Reference:
[1] Michaelis, 16th Int. Symp. on Flow Visualization, Okinawa, Japan, 2014
11
Tomo-PIV
Imaging Systems
Modular and upgradable Tomo-PIV solutions
LaVision provides a complete range of flexible and modular Tomo-PIV
systems and components for an integrated easy to use Tomographic
PIV solution for your application.
Tomo-PIV measurements are already possible with a two camera StereoPIV system. Each additional camera added to the system increases the
achievable spatial resolution exposing 3D flow structure in more detail.
Cameras
12
Components
3D Calibration Targets
13
3D Flame
Imaging
3D Flame structure and
turbulence-flame interaction
3D (Laser) Imaging in flames improves our understanding of combustion
processes and thus the realization of more efficient and cleaner combustion devices. The 3D distribution of flame radicals mainly located in the
reaction zones is visualized using 3D reconstruction of multiple camera
views and turbulence flame interaction in combination with Tomo-PIV
[1,2, TU-Darmstadt].
While eight 5.5 Mpixel Imager sCMOS cameras are used in parallel
to reconstruct the instantaneous 3D flame structure shown to the
right, time-averaged 3D flame imaging is possible with only one
camera collecting consecutively the flame emission from multiple
viewing angles.
References:
[1] Weinkauf et al., 17th Int. Symp. on Applications of Laser Techniques to Fluid Mechanics, Lisbon, 2014
[2] Weinkauf et al., Exp. Fluids 54, 2013
14
3D Spray
Imaging
Volumetric spray velocimetry
For many spray processes the knowledge about the spatial distribution
of the droplet velocities within the multiphase flow is of upmost importance especially for unsteady and highly dynamic sprays. For such spray
investigations only Tomo-PIV applied on spray droplets can measure the
3D spray structure of droplet motion in an instant.
15
Tomo-PIV
around Propellers
Propeller wake analysis
The knowledge of the mechanisms of wake instability behind rotor
systems, such as propellers, wind turbines or helicopter rotors, plays
an important role in many engineering applications, because of its
direct correlation to performance, vibrations, noise and structural
problems. Tomo-PIV represents a very effective tool for instant flow
based analysis as, for example, to study the three dimensional and
turbulent dynamics of the propeller tip vortices in the transition and
far fields.
camera 2
camera 1
water filled
prism
light volume
encoder
camera 3
water filled
prism
PTU
PC
camera 4
synchronizer
16
Tomo-PIV
in Engines
In-cylinder 3D flow fields
Knowledge of in-cylinder flow motion is essential for the design of modern
IC-engines exhibiting improved efficiencies and reduced emissions.
Four 1.4 Mpixel Imager intense cameras are detecting in a time-correlated
double exposure mode the movement of oil droplets seeded into the incylinder air. The measurement volume is illuminated with a double-pulse
Nd:YAG laser enabling Tomo-PIV measurements.
Exhaust (E)
Intake (I)
27
32
Volume
Optics
Cam
Reference:
[1] Baum et al., Flow Turbulence Combustion 92, 269-297, 2014
17
Cam
am 2
Spectra Physics
PIV - 400
E
Cam
Biomedical
Flows
3D blood flow visualization
Coronary artery diseases are the leading cause of death in the developed
world and are dependent on the 3D blood flow structure and shear wall
stress within the blood vessels. Tomo-PIV is successfully applied to investigate in vitro flow and wall shear stress in realistic coronary artery
models.
Normal artery
Liquid deposit
of plaque
no index matching
Narrowing of artery
This fluid is also filled in the test section surrounding the silicone
model to eliminate image distortion effects caused by refractive index changes in the different media.
fluorescent
particle image
18
Tomo-PIV
3D flow structure in a
human carotid artery
Courtesy of D. Obrist
University of Bern
Reference:
[1] Buchmann et al., 8th Int. Symp. PIV Melbourne, Australia, 2009
19
Volume (mm3)
Tracer particles
800
200 x 200 x 50
1 m oil droplets
200
120 x 120 x 60
50 m polyamid
20
34 x 30 x 19
1 m oil droplets
20
Helium bubbles
22.5
170 x 100 x 80
LaVisionUK Ltd
LaVision GmbH
Anna-Vandenhoeck-Ring 19
D-37081 Goettingen / Germany
E-Mail: [email protected]
www.lavision.com
Tel.: +49-(0)5 51-9004-0
Fax: +49-(0)551-9004-100
LaVision Inc.