Content uploaded by Marie Burckbuchler
Author content
All content in this area was uploaded by Marie Burckbuchler on Oct 12, 2021
Content may be subject to copyright.
13th International Symposium on Ultrasonic Doppler Methods for Fluid Mechanics and Fluid Engineering
Zürich, Switzerland (13.-15. June 2021)
Towards handheld UVP
Marie Burckbuchler1, Stéphane Fischer1
1Ubertone SAS, 14 rue du Brochet, 67300 Schiltigheim, France
Through its hardware size, the installation and handling of UVPs was not convenient. Also the setup complexity
limited the UVP technology to trained users or ultrasonic technology experts. In this extended abstract, we
present the world's first handheld UVP, which simplifies its use on large setups and moving systems. The UB-Lab
P has, through the presented results, proven its efficiency in hydraulics engineering studies on physical models,
especially with a moving carriage. The high sensitivity of the UB-Lab P enables measurement even in low
scatterers suspension conditions, such as low flow rates. The cloud.ubertone.eu processing and visualization tool
allows fast visualisation of raw measured data. It also provides processing scripts to prepare the data for
non-acoustics-experts, synchronising carriage positioning systems with velocity measurements, processing the
raw Doppler velocity measurements in multi-components vectors and filtering data.
Keywords: UVP, compact, autonomous, hydraulics, flow scanning
1. Introduction
As Ubertone’s first goal is to bring the UVP technology
to more users and applications, we follow several RTD
(research and technology development) axes. One axis
concerns the multiplication of measured velocity
components along a profile, while keeping a high
space-time resolution, which led to the UB-Lab 2C,
launched in 2018, a commercial ADVP (Acoustic
Doppler Velocity Profilers). Another axis is mobility and
easy installation with miniaturisation and autonomy.
While UVPs are usually of the size of PC towers,
Ubertone released its first compact laboratory UVP in
2011. Continuing in this direction, we launched last
september 2020 the UB-Lab P, the world's first handheld
UVP, a ready-to-use battery-powered instrument, with an
embedded web-interface accessed by wifi.
Apart from the convenience of having instruments of
small size and wireless that can be controlled with a
smartphone, these characteristics may be a strong
advantage when working on large setups, industrial plants
or on moving systems. This is the case in many physical
models used in hydraulic engineering [1] or on small
rivers [2] in which high resolution velocity measurements
are needed.
This presentation will focus on the UB-Lab P, with an
application example on a physical model of a river
section.
2. Material and Method
2.1 The UB-Lab P
The UB-Lab P is a handheld profiler equipped with a Wifi
connexion and an internal battery in a compact enclosure.
Some specifications are given in Table 1.
Table 1: Specifications of the UB-Lab P
Emitting frequency
0.025 to 3.6MHz
Spatial resolution
Down to 0.73 mm
Cells number
200
Transducers channels
2
Velocity accuracy
0.2 to 1 %
Sampling rate
up to 15 Hz
Trigger
IN/OUT
Communication
Wifi
Weight
0.2 kg
Consumption
2.5 W, on battery
A trigger ‘in’ or ‘out’ is available on the device. The input
mode has been used in this measurement campaign.
When a recording is started through the graphical user
interface, the input signal enables the storage of data.
Only blocks where the “trigger” signal is at high level (at
the beginning and the end of the block) are effectively
stored in the record file.
2.2 UVP Measurements
The UVP (Ultrasonic Velocity Profiler) technique has
been introduced to Fluid Mechanics by Takeda [3].
This technique based on coherent Doppler provides
velocity profiles measurements with a high spatial and
temporal resolution. An ultrasonic pulse is emitted in a
narrow beam and the particles, suspended in the flow,
scatter the pulse. The echoes of the particles are received
by the same transducer (monostatic mode) which allows
the observation of a profile composed of many
measurement cells distributed along the beam axis. The
signal is processed providing information of the velocity
component projected on the transducer’s axis.
Combining cells in two diverging beams allows getting
average two-components velocity profiles.
1 Towards handheld UVP
2
Figure 1: Combining two velocity projections on two
transducers to get the velocity vector orientation and module
The backscattered echo amplitude is recorded
simultaneously with the velocity measurement. The water
level (or of any other interface) can be evaluated by
observing a strong variation of the backscattered echo
amplitude gradient.
2.3 Setup
The flume is a physical model of a river with a hydraulic
structure and is situated in the hydraulics lab hall of the
CNR - Compagnie Nationale du Rhône, in Lyon, France.
The maximal depth in the flume is around 50 cm and flow
velocities vary from -10 to 50 cm/s.
The CNR was interested in the flow mapping over the
whole physical model. They defined a few specific
transects of interest.
Figure 2: Flume sketch. Hydraulic structure situated
downstream in the bottom left corner. Transects of interest
indicated with lines.
The flume was equipped with a programmable carriage,
recording xyz positions and able to manage a few sensors.
Figure 3: Carriage above the flume
2.4 Installation
The instrument’s small size, weight and wireless
communication make the installation easy.
As the access to the graphical user interface goes through
a common web browser and is compatible with a
smartphone, the use of this device is plug-and-play
simple.
The internal battery of the UB-Lab P has a 4 hours
autonomy. The use of an external power bank can
enhance the autonomy up to a few days.
The trigger input of the UB-Lab P was connected to the
carriage control unit.
Figure 4: UB-Lab P (within the green circle) installed on the
carriage
Two diverging transducers are used to get average
two-component velocity profiles along the vertical axis,
and a motor orients the transducer holder between U/W
or V/W plane.
3
Figure 5: Transducers holder beneath the carriage (within the
yellow circle in fig 4)
2.5 Procedure
The recording is started on the UB-Lab P, then the
carriage routine is started. Profiles indeed recorded by the
UB-Lab P are downloaded, after stopping the recording,
in a binary data file (.udt). Human readable .csv files can
be extracted from it when uploading it to the
cloud.ubertone.eu processing and visualization tool.
The carriage xyz positions are recorded simultaneously
and custom scripts on Ubertone’s cloud allow
synchronizing data from the carriage file with the
UB-Lab P data. We also adapted our interface detection
to obtain a robust and accurate measurement of the
riverbed position, using echo amplitude profiles.
In those scripts, the velocity data beyond the detected
interface are kept out and the data are filtered with the
Doppler signal-to-noise ratio (SNR) measured by the
UB-Lab P. U and W components are processed from the
monostatic projected velocities and the Doppler angles
given by the user. Other input parameters allow taking
into account environment aspects such as model scale, for
lab vs field measurement comparisons.
For the measurements presented here, two 3MHz
transducers were used, with a spatial resolution of 5mm.
The UB-Lab P switched periodically between both
transducers to alternate the measurements of both
monostatic components. Each cycle lasted about 1s.
3. Results and Discussion
3.1 Dynamic measurements
The first routine was to scan several transects and record
continuously during the displacement, the trigger was set
to low when transiting between transects to avoid
recording.
In the following figures, in each group of three graphs,
the first above represents the U component profiles, the
central the W component profiles and the one below the
distance between the transducer and the flume bed. The
y-axis is the depth and the x-axis is the resulting profile
number obtained along the transect.
Figure 6: Processed dynamic transects acquisition
3.2 Static measurements
As for some discharge conditions, the signal was weaker,
scatterers settling faster, another routine was to measure
in fixed positions of the transect for a few seconds.
Another script developed by Ubertone on the cloud
application then averages the data collected in each fixed
position for a better spatial cover.
Figure 7: Processed static-positions-wise transect acquisition
1 Towards handheld UVP
4
In Fig. 7, the first display shows the measurements
processed by the dynamic script. The second shows the
result after averaging per fixed position.
6. Summary
Using a handheld UVP device can be helpful in
mechanical design of setups, especially where the weight
of the moving carriage is critical. The wireless
communication and plug-and-play simplicity enables
easy on-the-spot UVP measurements.
The next step in handyness improvement will be a
wireless trigger management.
As shown in this abstract, the UB-Lab P’s high sensitivity
results in good quality velocity profile measurements
even when few scatterers are in suspension. The Doppler
SNR filtering improves the quality of the average
velocity.
The echo signal is used to detect the bed and prepare the
data for post-processing analysis.
The cloud.ubertone.eu processing and visualization web
tool easens the work for non-acoustics-experts, providing
processing scripts. The data is prepared for further
analysis: synchronisation of carriage positions with
velocity measurements, processing of raw Doppler
velocity measurements in multi-components vectors and
data filtering.
Now for hydraulics studies, the following development
will be to measure 3 components by adding transducer
channels.
References
[1] De Cesare G, et al.: River intake and desander efficiency
testing on a physical model using UVP and LSPIV, ISUD 7
(2010).
[2] Burckbuchler M & Fischer S: A miniature uvp hardware
applied to environmental monitoring, HydroSenSoft (2017).
[3] Takeda Y: Measurement of velocity profile of mercury flow
by ultrasound Doppler shift method, Nucl. Technol. 79
(1987), 120-124.
5
