XVIIèmes Journées Nationales Génie Côtier – Génie Civil

Chatou, 2022
DOI:10.5150/jngcgc.2022.102 © Editions Paralia CFL
disponible en ligne – http://www.paralia.fr – available online

Field velocimetry for sediment transport studies

in water and air

Callum GRAY 1, Christy SWANN 2, Joe CALANTONI 2, Blake LANDRY 2,

Carlo ZAMALLOA 2

1. LaVision Inc, Ypsilanti, MI, USA.

callum@lavisioninc.com
2. Ocean Sciences Division, Naval Research Laboratory, Stennis Space Center,
MS, USA.
christy.swann@nrlssc.navy.mil; joe.calantoni@nrlssc.navy.mil;
Blake.Landry@nrlssc.navy.mil; Carlo.Zunigazamalloa.ctr.pe@nrlssc.navy.mil

Abstract:
Rates of sediment transport predicted by models differ significantly from field
observations in both aeolian and underwater environments. Refinement of predictive
models requires high quality ground truth data to discern if the discrepancy between
models and field observations lie in a given model’s ability to properly represent the
physics of motion, or if our field observations are inadequately sampling the sediment
transport process. Identifying the sources of error can only partially be achieved in a
laboratory wind tunnels or wave flumes. The limited length and scale of turbulence
recreated in wind tunnels do not follow those of natural boundary layer conditions that
exist on a windy beach. Similarly, wave amplitudes and spectral content resulting in
seabed sediment transport can only be recreated in limited scope in a wave flume or
oscillating water column. Ruggedised point measurement tools are available for
measuring fluid velocities in a field environment but are of limited spatial and temporal
resolution. Here we describe the adaptation of two high resolution full field velocimetry
techniques that fully transition laboratory methods to quantify sediment transport by wind
and waves for use in field environments. To measure subaqueous sediment transport, we
developed a Stereo Time Resolved Particle Image Velocimetry (SPIV) system that
utilizes submersible enclosures for remote operation in an undersea environment at depths
of up to 15 meters for a period of several weeks. The system was deployed off the coast
of O’ahu near the Makai Research pier. To measure sediment transport by wind in a
natural boundary layer, we developed a Time Resolved Volumetric Velocimetry system
based on Shake-the-Box 4D Particle Tracking (STBPT). We deployed this system on a
secure beach at NASA Wallops Flight Facility in Wallops Island, Virginia. Adapting
valuable and sensitive equipment to measure sediment transport by waves and wind in an
outdoor environment presents substantial technical and logistical challenges. This paper
will describe the two systems mentioned above including design considerations, technical

1019
Lecture spéciale

innovations, experiences gleaned from the measurement campaigns and examples of data
obtained.

Keywords:
Sediment transport, Aeolian, Maritime hydraulics, Stereo PIV, Shake-The-Box particle
tracking, Maritime works, Coastal environment, Coastal ecosystems.

1. Introduction
There are certain advantages and disadvantages associated with making experimental
measurements in the field versus the coastal Engineering laboratory. The first and most
obvious is that in the field you experience “real” phenomena which incorporates all the
subtle nuances of the real world that cannot be easily or sometimes ever reproduced in
the laboratory. At local scales (near field) it is typically straight forward to reproduce
magnitudes of velocity that are of the same or very similar magnitudes of velocity
observed in the field. However, what cannot be reproduced is the “far field” physics. We
cannot, for example, simulate the modulation of energy from short period waves to
infragravity waves with periods of the order of minutes. Additionally, in the lab
everything is “unidirectional” whereas in the field directions are oscillating all the time
and there is a directional spectrum. Also, complex sedimentology beyond a mean grain
size and local geology cannot be reproduced in the lab. On the other hand, the range of
conditions that can be sampled in the field is limited by the prevailing weather occurring
across the time window available. Also, one cannot rely on conditions remaining stable
to collect statistically convergent data and real-world conditions may be much more
complex than a deliberately simplified lab experiment. For these reasons it can be much
more difficult to develop “theoretical” or more easily generalizable knowledge from field
measurements. Lab based experiments are much better suited to theoretical studies and
for isolating one or two processes to determine their respective impact on the outcome.
Lab experiments are less expensive to perform, can be repeated many times and are better
suited to numerical modelling to close the picture. In the lab you can essentially know
everything. All details and boundaries etc whereas in the field it is impossible to know
everything that may influence your results. In our experience working in the field is
usually necessary when you have a very specific application that is be explored or
developed.
The rest of this paper describes two separate field deployed velocimetry systems that have
been adapted directly from laboratory-based equipment originally implemented in the
costal engineering lab of the Naval Research Laboratory at the Stennis Space Center.
The first system to be implemented was a Stereo Time Resolved PIV system (ADRIAN
& WESTERWEEL, 2011; RAFFEL et al., 1998) designed to operate at sea depths of 10-
15 meters and at distances from shore of up to 1.5km. The system is self-contained
requiring only a 300V DC power source and optical fibre to communicate to shore. The

1020
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

cameras, light source and control system were protected from the undersea environment
by mounting into heavy duty enclosures with windows which themselves were mounted
under a rotatable plate which was supported under a “quadpod”. The orientation of the
stereo light sheet could be selected to suit the prevailing wave direction. The system was
deployed off the coast of O’ahu near the Makai Research pier during the summer of 2018.
The design and implementation of the system started in 2016.
The second project that brought a traditionally laboratory technique into the field was a
volumetric velocimetry system based on the Shake-The-Box 4D particle Tracking method
(SCHANZ et al., 2016). The system was deployed on a secure beach on Wallops Island
Virginia in October 2021 and designed to measure with high spatial and temporal
resolution the turbulent boundary layer across the beach with the aim of understanding
the relationship between the aerodynamic flow field and the “streamers” of saltating sand.
Here again the system was designed to be resistant to the environment and allowing it to
be utilised for extended periods of time (weeks).

2. Implementations

2.1 Underwater Time Resolved Stereo PIV System

The motivation for this project is to provide high-fidelity, laboratory-type measurements
of the seabed 2D 3C velocity field and sediment motion across a representative range of
real ocean conditions. The intention is to gain a more complete understanding of how
near bed turbulent fluctuations initiate the mobility and burial of unexploded ordinance
(UXO) and to calibrate and validate predictive models. The Imaging System for Littoral
Environments (ISLE) was developed at the US Naval Research Laboratory and is a
ruggedised Time Resolved Stereo Particle Image Velocimetry mounted under a quadpod
along with other optical, electronic, and acoustic sensors. To facilitate long term
deployment of up to one month ISLE was cabled to, and operated from, the shore. Repeat
dive operations were conducted to provide in-situ maintenance. This includes mitigation
of biofouling and to position and reposition different types of UXO in the test area.
Particle Image Velocimetry (PIV) is a well-known technique for flow mapping and has
been used extensively in the lab for this purpose. The principles of PIV are well
documented (ADRIAN & WESTERWEEL, 2011; RAFFEL et al., 1998) and will not be
covered here. The requirement for 3 components of velocity requires Stereo PIV (SPIV)
necessitating the use of two cameras. Several wave periods of >10 seconds were to be
acquired at a frequency sufficiently high to capture the characteristic timescales of the
flow across a range of real ocean conditions. High Speed CMOS cameras were used to
capture the stereo images and a scanned CW laser was used to illuminate the region of
interest. The combination of high frame rate CMOS cameras and continuous scanning
permits acquisition frequencies up to 3.5 kHz though in practice lower frequencies were
sufficient to capture the temporal features of interest. The desired field of view was 1.2

1021
Lecture spéciale

meters wide and approx. 50cm high. Wide enough to simultaneously image several sand
ripples and high enough to capture the flow structures from the interaction between bed
ripples and wave/current. Another important consideration is that the submerged
components should not impede or modify the flow field being studied and should have a
small profile to minimise drag on the structure. The enclosures were designed to be
deployed for extended periods of time in sea water to resist biofouling and be capable of
tolerating a range of temperatures from a hot ship deck to cold under sea conditions.
The concept that was developed is based on the use of four separate watertight enclosures
containing the cameras, the light sheet generator and the control enclosure that contains
an embedded PC, Timing Unit and power supply. These four enclosures are rigidly
mounted to a circular metal plate which itself is mounted under a “Quadpod”. A schematic
of the whole system is shown below in figure 1.

Figure 1. Schematic of the TR SPIV system mounted to orientation table and quadpod.

As can be seen the two camera enclosures are mounted looking down at the measurement
area and inwards providing a stereo angle. The scanning light sheet enclosure is mounted
vertically projecting the laser sheet downwards to the seabed. The “control” enclosure is
mounted on the top shelf and provides power, timing and a data interface to the camera
and laser enclosures. The positioning of the camera and laser enclosures are high enough
above the bed to avoid modifying the flow being measured.

1022
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

2.1.1 Camera Modules

All the enclosures were designed and built by Proteus Technologies from Slidell, LA and
were of similar design to reduce complexity of manufacture.
Enclosure material: Machined aluminium bronze alloy C955. This material is of the order
90% copper which is toxic to living organisms, and this reduces biofouling which is a
problem when submerged in sea water for extended periods of time. The mechanical
strength of the enclosures without a glass window would have permitted use at depths of
1000 meters. However, the pressure rating of the housings with the quartz window
recessed into forward end caps and glued was 50 meters with failure at 150 meters. This
far exceeds the target depths of 10-15 meters. The bronze material also has reasonable
thermal properties to conduct heat.
Enclosure Design: The smallest possible cylinder diameter was used to reduce drag. The
fore and aft end caps were designed to incorporate double o Rings. This is based on a past
design and used many times to submerge other kinds of instrumentation (acoustic) to
significant ocean depths.
The cameras (VEO340L Vision Research) were fitted with 50mm f1.8 Canon lenses with
remote focus and aperture control. The Scheimpflug angle was calculated for the
orientation of the cameras relative to the light sheet. A “compound” Scheimpflug angle
is used to tilt the focal plane of the camera looking inward and downwards. This
compound angle is used when the stereo viewing angle is not just in the horizontal plane
bit also in the vertical. Due to the remote aperture and focus it was not possible to
incorporate an adjustable Schiempflug adapter and so a fixed angle device was designed
that could fit between camera and lens without increasing the lens sensor distance and so
the full range of focussing including infinity was preserved, figure 2.

Figure 2. Fixed angle Schiempflug adapter.

1023
Lecture spéciale

The non-window end caps all had “pass-through” connectors that completely seal the
enclosures but permit the required lines for power, triggering and data. The connectors
were Subconn underwater mate able connectors apart from the fibre umbilical going into
the control enclosure and that was GISMA fibre/copper connector, figure 3.
To avoid condensation of trapped humidity within the enclosures when sealed and put
into cold water the enclosures were evacuated to low pressure and then backfilled with
dry nitrogen. Humidity sensors were installed in all encloses to detect water ingress.

Figure 3. Camera enclosure showing the cylinder, end caps with pass through
connectors and window.

2.1.2 Laser Module

As mentioned, the laser sheet system was based on a scanned CW laser beam from a CW
YAG laser module outputting >6 watts at 532nm. Laboratory based PIV systems have the
benefit of laboratory infrastructure specifically including access to ready cooling. Most
PIV systems utilize a pair of flashlamp pumped, q-switched, frequency doubled laser
cavities that typically generate 100-200mJ of energy in pulses of < 10ns duration. These
kinds of lasers are efficient but require a considerable amount of active cooling which is
provided in the lab by a water to air or water to water chiller. Additionally, the need for
active water cooling adds bulk to the laser head. The solution to this is to use a laser that
is compact, has only one cavity, does not require active cooling and coincidentally is a
fraction of the cost of a double cavity pulsed PIV laser. We used a 6.51W (measured)
DPSS CW YAG laser. This is possible for relatively low velocity flows by using the
scanning beam method (GRAY et al., 1991). The scanning beam light sheet was

1024
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

developed for wave studies and is known to work for significant fields of view up to 1
meter so long as the maximum velocities are in the range of a few m/s. This approach
would not be suitable for higher velocities or PIV in air but in this case permits a very
compact light sheet with repetition rates up to 3000 scans per second. In the actual trial
of the system scan rates < 1kHz were used and so well within the operating range of the
system, figure 4.

Figure 4. Laser sheet enclosure showing laser, 20 facet scanning mirror and driver.

2.1.3 Control Module

The control module contains a host PC (embedded compact PC), a timing unit (LaVision
PTU X) which had been extracted from its enclosure, a “smart” power supply that takes
300V DC and converts to 24V, 12V, 8V to suit the requirements of the cameras, remote
lens control, humidity sensors, laser, scanner, PTU X etc. The power supply has a
controller that has a software interface that is used to control and monitor the power
supply and ensures that the laser is not turned on until the scanner mirror is spinning. This
avoids burning the mirror surface. The controller also provides feedback of the internal
temperature of the enclosures from a thermocouple in each. Even though the heat
generated by the CW laser was not great it was automatically switched off between runs
as extended use would trip the lasers own temperature interlock. To increase passive
cooling of the laser the base of the laser was connected to a metal bridge to the inside
surface of the enclosure. Thermal conductivity was enhanced by using a thermal paste
between laser base and heatsink, figure 5.

1025
Lecture spéciale

Figure 5. Control enclosure showing embedded PC, Timing unit (PTU X), power supply
and control electronics board.

2.2 Deployment
The system was assembled at Stennis Space Centre in Mississippi where the Naval
Research Laboratory have facilities and subjected to several tests concluding with a full
immersion test. Once the integrity of the cannisters was confirmed the system was
initialized and started acquiring images. The outdoor tank shown in figure 6 was seeded
with 20-micron polyamide spheres and an induced circulation flow in the tank measured
after performing a stereo calibration. This test confirmed that cameras, laser, and scanner
were all operating as expected without overheating and that the embedded PC could be
accessed by remote desktop via the 1500 meters fibre optic ethernet cable.

Figure 6. Immersion test of system at Stennis Space Center.

1026
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

Following these tests made in March 2018 the system was disassembled and packed for
shipment to the Makai Research Pier on the island of O’ahu, figure 7. In July 2018 the
system was assembled under cover of the pier building. An electronics test was performed
to ensure everything is operating correctly tested and then lowered into the sea through
the removable floor of the pier.

Figure 7. Makai Research Pier O’ahu.

The quadpod was fitted with several flotation balloons that rendered the quadpod
neutrally buoyant in the sea water. With the quadpod suspended under the pier and
maintained in position by the pier crane it was left for several hours with the cameras and
the laser operated periodically. The humidity sensors in the cannisters were monitored to
verify that everything was watertight. A minor leak in one of camera cannister was
detected after some time. The quadpod was lifted out of the water and the suspect
cannister removed inspected, figure 8. A crack had developed in the glue that held the
quartz window in place. This was repaired and the same “dunk test” performed and this
time everything checked out. When no water ingress was detected after 8 hours in the
water the system was left submerged under the pier overnight but monitored by a
researcher from the remote desktop overnight just in case a leak developed again.
The next morning a team of divers manoeuvred the quadpod out from under the pier and
out into open water to the test location. The Makai research pier was selected because it
was a secure location with facilities well adapted to these kinds of activities. However,
the pier is also located near to protected sea life and because we were using a class 4 laser

1027
Lecture spéciale

it was a requirement that the quadpod be fitted with 360-degree colour cameras that could
be used to detect any protected species that may come close to the test site, figure 9.

Figure 8. The quadpod being lowered into the sea through the removable floor of the
pier.

At this point we were able to start operating the system. A piece of decommissioned
ordinance (a shell) was placed under the quadpod in a position that the laser sheet
intersected. Divers positioned a 3D calibration target in the plane of the laser sheet and
the system was calibrated. In a lab-based PIV experiment seed particles will be added to
the fluid. In air an aerosol of oil or helium filled soap bubbles will be used to provide
neutrally buoyant tracers. In water seed particles made from glass or polyamide are
commonly used. In many cases the particles will be fluorescent to be used with high pass
filters on the cameras to exclusively image the particles and improve image quality for
PIV processing. However, in the sea it is not practical to artificially seed the flow. Sea
water contains a relatively high density of naturally occurring suspended particles which
we anticipated would be sufficient for PIV. Operating the system while suspended under
the Makai pier it was confirmed that a sufficient number density of particles was present
and visible. The laser scanner was operated at scan rate of 330 scans per second. This

1028
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

gives a PIV delta T of just over 3ms and resulted in particle displacements of around 8-
10 pixels.

Figure 9. Camera and laser cannisters mounted under the quadpod on rotation table.

Stereo PIV Images were acquired to the memory of the CMOS cameras and when the
18Gigabytes of memory in each camera was full the sequences were downloaded to a
networked disk for processing offline. The TR SPIV image acquisition and processing
was performed from DaVis 10 with Stereo PIV processing module (LaVision GmbH).
Image quality was considered good with contrast of nearly 1000 counts between particle
images and background. Measurements taken during the middle of the day when the
overhead sun was brightest tended to lose some contrast. Band pass filters on the camera
lenses (532nm +/- 10ns, 95% transmission) were used to reduce most of the background
ambient light. Most conditions of interest tended to be in the evening or at night in which
case the background light was no issue.
Figure 10 below shows an example of the velocimetry data measured with the system.
On the right shows the velocity field over a pair of sand ripples with vorticity as the colour
code. The legend on the right shows the relationship between colour and vorticity
magnitude

1029
Lecture spéciale

Figure 10. Example velocity plot acquired from ISLE showing vorticity as colour and
on left a temporal trace taken from one point in the imaged field of view.

2.3 Aeolian Volumetric Velocimetry System

Existing aeolian transport models often fail in field environments. The discrepancy
between models and prediction has been attributed to inadequate field measurements and
uncertainty in the fundamental processes (turbulence) driving sand transport. The main
challenges of measured and modelled aeolian transport include:
1) Coarse resolution measurements relative to the fluid -grain scale physics
2) An inadequate understanding of the fundamental physical relationship between
turbulence and sand transport
3) The inability of aeolian transport models, derived from wind tunnel observations, to
simulate natural boundary layer processes at the appropriate field scales (mm-cm)
The Field based Particle Tracking Velocimetry (F-PTV) system was developed at the
Naval Research Laboratory (NRL) to address these challenges. Prior measurement
techniques integrate over sample volumes on the order of cm to tens of cm and often
cannot be placed in the saltation cloud. For example, ultrasonic anemometer. Or, in the
case of cup anemometer, has a response time that precludes sampling turbulence at the
relevant timescales. Also, these kinds of instruments measure turbulence at a point and as
such as incapable of adequately sampling the spatial and temporal fluctuations and lags
that cause turbulence induced transport.
The F-PTV system is based on components that are previously exclusively used in a
laboratory and acquired images processed using the LaVision Shake-The-Box (STB)
particle tracking algorithm. A detailed description of the STB method will not be given
here (SCHANZ et al., 2016) though a general overview is helpful. Neutrally buoyant
tracers are injected into the flow and illuminated by a pulsed light source to render the
flow visible as with PIV. Unlike conventional PIV though a volume rather than a plane
is illuminated. The light source pulses repeatedly and 4 high frame rate cameras
synchronised with the pulsed illumination capture a sequence of images of the scattered
light. The 4 cameras are calibrated using a flat dot target so that positions in the

1030
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

measurement volume can be accurately related back to positions on each of the 4 camera
sensors. The camera mapping functions that are generated from the calibration images
are further refined using a technique called Volume Self Calibration (WIENEKE, 2008)
which uses the experiments data images to “tune” the mappings so that particle locations
can be determined to better than 0.1 pixel. The Shake the Box (SCHANZ et al., 2016)
algorithm uses the data images to triangulate particle positions in the real world from the
2D pixel images. The 3D coordinates of found particles are iteratively processed into
trajectories which may then be converted from Lagrangian trajectories into a Eulerian
grid of vector values. A simple but fast binning method can be used to re-grid the data, or
a more sophisticated approach uses a 4D solver to come up with a solution constrained
by the known trajectories. This is called Fine Scale Reconstruction (JEON et al., 2022)
and is based on a solver called VIC#.

Figure 11. Schematic of the field PTV system showing Helium Filled Soap Bubble
Generator (HSFBG), gantry with optical fibre fed volume optics and mounted cameras.

Figure 11 shows a schematic of the F-PTV system. Artificial seeding is injected into the
flow upwind of the measurement system. The tracers drift through the measurement
station which is a gantry from which the illumination is projected down to the beach The
light scattered by the tracers is imaged by an array of 4 high speed cameras looking
upwind at the illuminated seeding plume. The system is not completely non-invasive but
minimally so.
The seeding system used was a LaVision Helium Filled Soap Bubble Generator (HFSBG)
(FALEIROS et al., 2019) which generates 230-micron helium filled soap bubbles. At this
specific size they are neutrally buoyant and have the primary advantage for this
application that they are large and scatter approximately 10,000 X more light intensity
than the aerosol commonly used in wind tunnels. The HFSBG consists of a controller
with a soap reservoir which is connected to Helium and Nitrogen tanks. The controller
ensures the correct pressure and flow rate of soap, He and N2 is maintained to generate

1031
Lecture spéciale

the bubbles and 40,000 are produced per second from each nozzle. There are 20 nozzles
per linear nozzle array (LNA) and as many as 10 linear nozzle arrays can be supported
from one controller, figure 12. For the tests at Wallops Island 3 LNA’s were used. A
custom case was designed to protect the controller from sand and salt. To avoid the LNA
from altering the flow being measured the LNAs were laid sideways into wooden boxes
which were lowered into a trench dug into the sand up wind of the gantries.

Figure 12. Helium Filled Soap Bubbles being generated from Linear Nozzle Array.

Pulsed illumination is provided by Diode Pumped solid state YLF laser (Photonics
Industries DM Dual head YLF). The laser is a sensitive instrument and is protected from
the elements by mounting into a weatherproof cabinet along with the other sensitive
components such as the timing unit (LaVision PTU X HS) and the fibre optic beam launch
optics. The cabinet itself was mounted onto the side of the gantry (see figure 13) that
supports the volume optics and the other pieces of instrumentation such as cup
anemometers, sonic anemometers, moisture probes, saltation sensors and saltation traps.
The instrument gantries are shown below in figure 13. The main gantry mounts the laser
volume optics and the other probes mentioned. The camera gantry provides a rigid
platform for the cameras. The cameras are protected from the sand, salt, and rain by
camera enclosures like those used for outside video surveillance. Extra vents were
installed in the back of these enclosures to enhance cooling of the cameras. Each camera

1032
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

was fitted with the same remote focus and aperture controllers as used in the underwater
PIV system described in the first section. This permits control of focus and aperture on
the lenses of the enclosed cameras.

Figure 13. Schematic of the Field PTV System layout.

The cameras are VEO340 L CMOS high frame rate cameras from Vision Research. A 1-
meter x 1-meter calibration target was mounted along the mid-section of the illuminated
volume and used to generate camera mapping functions for each of the 4 cameras. Once
the standard calibration was captured then a test series of images of helium bubbles was
captured and used to perform volume self-calibration. Volume Self-Cal is used to enhance
the accuracy of the mapping functions down to below 0.1 pixel required for reliable
detection and tracking of the 3D bubble trajectories.

1033
Lecture spéciale

Figure 14. Picture of the F-PTV system on the beach at Wallops Island. Weatherproof
cabinet shown on right attached to the instrumentation gantry.

The F-PTV system was deployed at NASA Wallops Flight Facility on the eastern shore
of Virginia from September 29th through October 28th, see figure 14. The instrument
gantry was located approximately 6 meters shoreward of the of the base of the dune and
20 meters from the shoreline. Initially, the system was operated to test for the optimal
locations of the HFSBG LNAs to distribute the helium bubbles. LNA locations turned
out to be a critical parameter for successful seeding of the measurement volume. Changes
in wind direction sometimes required relocation of the LNAs and if operated at night
required some choreography. Future measurement campaigns will benefit from having
more LNAs and having them more closely spaced and across an arc upwind of the
measurement station. With the limited LNAs available at the time many of the captured
image sequences had inhomogeneous seeding giving dense results in some location and
times but infrequently uniformly seeding the whole imaged volume. A range of
conditions were sampled and some of those required operating at night. Data captured at
night-time had the advantage that less image pre-processing was required to remove the
bright background scene captured during the day. Work is ongoing to process all the data
captured and to contextualize it with the other data sets sampled at the same time. An
example of the trajectories and the remapped Eulerian data is shown in figures 15 and 16.

1034
 XVIIèmes Journées Nationales Génie Côtier – Génie Civil
Chatou 2022

Figure 15. Example map showing trajectories across much of a 1 m x 1m x 40cm

volume.

Figure 16. Lagrangian trajectories remapped to Eulerian grid showing central plane
within the measured volume.

3. Conclusions
The two examples of field velocimetry systems show that laboratory techniques such as
PIV and 4D PTV can be taken into the field and implemented to obtain laboratory like
results. The trials performed in Hawaii and Virginia will serve as templates for future
campaigns. Insights obtained from these initial trials will help guide future field exercises

1035
Lecture spéciale

both in terms of optimising the measurement systems as well as guiding what conditions
are of prime interest.

4. References
ADRIAN R.J., WESTERWEEL J. (2011), Particle image velocimetry, Cambridge
University Press, New York, 558p. ISBN:9780521440080 and 0521440084
RAFFEL M., WILLERT C.E., KOMPENHANS J. (1998). Particle image velocimetry –
A practical guide, Springer, Berlin, 255p. https://doi.org/10.1007/978-3-662-03637-2
GRAY C., GREATED C.A., MCCLUSKEY D.R., EASSON W.J. (1991). An analysis of
the scanning beam illumination system, Meas. Sci. and Techno., vol. 2, no 8, pp.717-724.
https://iopscience.iop.org/article/10.1088/0957-0233/2/8/003
SCHANZ D., GESEMANN S., SCHRODER A. (2016). Shake-The-Box: Lagrangian
particle tracking at high particle image densities, Exp Fluids, 57:70,
https://doi.org/10.1007/s00348-016-2157-1
WIENEKE B. (2008). Volume self-calibration for stereo PIV and tomographic PIV, Exp
Fluids, 45:549–556. https://doi.org/10.1007/s00348-008-0521-5
JEON Y.J., MULLER M., MICHAELIS D. (2022). Fine scale reconstruction (VIC#) by
implementing additional constraints and coarse-grid approximation into VIC+, (2022),
Exp Fluids, Vol. 63, Issue 4. https://doi.org/10.1007/s00348-022-03422-9
FALEIROS D.E., TUINSTRA M., SCIACCHITANNO A., SCARANO F., (2019).
Generation and control of helium-filled soap bubbles for PIV, Exp. Fluids, Vol. 60, Issue
3. https://doi.org/10.1007/s00348-019-2687-4

1036