-
1192 IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND
FREQUENCY CONTROL, VOL. 67, NO. 6, JUNE 2020
Transmission of Images With Ultrasonic ElasticShear Waves on a
Metallic Pipe Using
Amplitude Shift Keying ProtocolAlexander Heifetz , Dmitry
Shribak, Xin Huang, Boyang Wang, Jafar Saniie, Life Fellow,
IEEE,
Jacey Young, Sasan Bakhtiari, Senior Member, IEEE, and Richard
B. Vilim
Abstract— Transmission of information using ultrasonicelastic
waves on existing metallic pipes provides an alter-native
communication option across physical barriers ina highly
partitioned industrial complex, such as a nuclearfacility. This
work investigates the feasibility of the trans-mission of digital
images over metallic pipes. Ultrasoniccommunication systems for
transmission of images ona nuclear-grade stainless steel pipe were
assembled forbench-scale demonstration. Information carriers in
this sys-tem are refracted shear waves transmitted and receivedwith
piezoelectric transducers (PZTs) operating at 2-MHznominal
frequency. The refraction and propagation of ultra-sonic shear
waves were modeled with COMSOL software.An amplitude shift keying
(ASK) communication protocolfor image transmission was developed
and implemented inthe GNURadio software-defined radio (SDR)
environment.Digital information was converted to analog ultrasonic
sig-nals using Red Pitaya electronic boards. The performanceof the
ASK protocol is evaluated at the output of every blockin the
GNURadio program by monitoring the transmission ofselect
characters. Using the ASK communication protocol,the transmission
of the 32-KB image was demonstrated at2-Kbps bitrate across
6-ft-long stainless steel pipe. Prelim-inary evaluation of
ultrasonic communication on the pipingof a nuclear facility, such
as signal transmission on bentpipes, was performed with COMSOL
computer simulations.
Index Terms— Amplitude shift keying (ASK), digitalimages,
ultrasonic transducers.
Manuscript received November 27, 2019; accepted January 16,
2020.Date of publication January 23, 2020; date of current version
May 26,2020. This work was supported by the U.S. Department of
Energy,Nuclear Energy Enabling Technology (NEET) Advanced Sensors
andInstrumentation (ASI) program under Contract
DE-AC02-06CH11357.(Corresponding author: Alexander Heifetz.)
Alexander Heifetz, Sasan Bakhtiari, and Richard B. Vilim are
with theNuclear Science and Engineering Division, Argonne National
Laboratory,Argonne, IL 60439 USA (e-mail: [email protected];
[email protected];[email protected]).
Dmitry Shribak was with the Nuclear Science and Engineering
Division,Argonne National Laboratory, Argonne, IL 60439 USA, and
also withthe Physics Department, University of Chicago, Chicago, IL
60637 USA.He is now with the Department of Electrical Engineering,
Georgia Insti-tute of Technology, Atlanta, GA 30332 USA (e-mail:
[email protected];[email protected]).
Xin Huang, Boyang Wang, and Jafar Saniie are with the
Electricaland Computer Engineering Department, Illinois Institute
of Technology,Chicago, IL 60616 USA (e-mail:
[email protected]; [email protected];
[email protected]).
Jacey Young is with the Physics Department, St. Norbert
College,De Pere, WI 54115 USA (mail: [email protected]).
Digital Object Identifier 10.1109/TUFFC.2020.2968891
I. INTRODUCTION
TRANSMISSION of information using ultrasonic wavesas information
carriers on pipes provides an alterna-tive communication option for
scenarios when conventionalwired or wireless communications are
ineffective or disabled.For example, using existing pipes as
communication channelsin a nuclear facility provides an option to
transmit informationto hard-to-reach places and across physical
barriers in thepost-accident scenario [1]. The main components of a
nuclearfacility are isolated from the outside world with
3–4-ft-thickconcrete walls of a containment building. In most of
thedesigns, the concrete walls of a containment building are
alsoplated with a metallic liner. A barrier of this type blocks
awireless RF communication channel [2]. Existing penetrationsof the
containment building wall consist of specially designedtunnels for
heat exchanger pipes, which deliver water fromand back to the
ambient reservoir [3]. The tunnels are sealedwith metallic plates
on both ends, which prevents the insertionof any electrical or
fiber optics communication cables. Themetallic pipes are not in
physical contact with concrete ofthe tunnel, which allows for the
propagation of elastic waveson pipes for sufficient distance to
traverse the concrete wallbarrier [4]. Because mounting ultrasonic
transducers on pipesinvolves minimal hardware modifications, such a
communica-tion system would be compliant with requirements of
nuclearfacility operations, which are subject to strict
regulations.In addition, this approach provides a degree of
physical cyber-security and accident resilience since the channel
consisting ofa metallic pipe is difficult to sever, compared to
conventionalcommunication cables [5].
In general, ultrasonic information transmission involveseither
free space (acoustic waves in fluids in solids) or guidedwave
(elastic waves in solids) communications. The majorityof work in
free space ultrasonic transmission is directed towardunderwater
acoustic communications [6], [7], but communica-tion through air
[8], [9] and metallic media have been studiedrecently [10], [11].
Guided-wave ultrasonic communicationon pipes combines elements of
communication theory withultrasonic transducers and wave
propagation in solids, whichare traditionally investigated in the
context of nondestruc-tive testing [12], [13]. Prior work on
ultrasonic information
0885-3010 © 2020 IEEE. Personal use is permitted, but
republication/redistribution requires IEEE permission.See
ht.tps://ww.w.ieee.org/publications/rights/index.html for more
information.
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
https://orcid.org/0000-0002-8891-9323
-
HEIFETZ et al.: TRANSMISSION OF IMAGES WITH ULTRASONIC ELASTIC
SHEAR WAVES 1193
transmission with guided elastic waves on metallic pipes
hasconsidered energy-efficient low bitrate (∼100 bps)
communi-cation using ON/OFF keying (OOK) modulation and chirpedOOK
to mitigate frequency selectivity of the channel [14],and explored
several approaches to modulation/demodulationusing time-reversal
position modulation [15], and cyclic fre-quency shifting [16]. In
addition, using different types ofultrasonic transducers, such as
the electromagnetic acoustictransducers (EMAT), for communications
using elastic guidedwaves on plates has been studied recently [17],
[18].
In this article, we investigate the ultrasonic transmissionof a
large volume of data, such as images, at a high bitrate, along the
metallic nuclear-grade pipes. Because of thelimited bandwidth of
ultrasonic transducers, amplitude shiftkeying (ASK) modulation,
which is one of the realizationsof OOK protocol, was used for
information transmission.A preliminary description of an ultrasonic
communicationsystem on pipes was presented in [5]. The
communicationscheme is implemented using GNURadio
software-definedradio (SDR) environment. Recent studies
investigated SDR-based communications using longitudinal wave
transmissionthrough solids and guided elastic waves communication
onplates [19]–[21]. Transmission of small volumes of data, suchas
text, has been demonstrated [21]. This article contains adetailed
analysis of the image transmission protocol developedin the
GNURadio SDR environment and demonstration of thecommunication
protocol performance in transmitting an imagewith ultrasonic shear
waves on a pipe. Images were transmittedalong the metallic pipe
using ultrasonic carrier frequencyelastic shear waves at 2-Kbps
bitrate. The analysis presentedin this article allows us to
evaluate the performance of theultrasonic communication system and
investigate strategies forfurther improvement.
This article is organized as follows. Section II discussesthe
hardware and software of the ultrasonic communicationsystem.
Section III presents the analysis of the communicationprotocol by
monitoring transmission of a simple messagethrough the
communication system. Section IV briefly dis-cusses potential
challenges for deployment of the ultrasoniccommunication in a
nuclear facility. Section V contains thesummary and
conclusions.
II. ULTRASONIC COMMUNICATION SYSTEM DESIGNA. Ultrasonic
Communication System Hardware
A literature review of common reactor designs has identifieda
chemical volume and control system (CVCS) water carryingpipe, which
penetrates the reactor containment building wallin several reactor
designs, as a viable conduit for communi-cations [3]. A laboratory
bench-scale system consisting of anuclear-grade CVCS-like stainless
steel pipe and ultrasonicpiezoelectric transducers (PZTs) was
assembled for a pre-liminary communication system analysis. The
pipe used inthe bench-scale study is a 6-ft-long schedule 160
stainlesssteel 304 L with 2.375-in outer diameter and 0.344-in
wallthickness.
This study used commercial paintbrush PZTs operatingat a nominal
frequency of 2 MHz. While other types of
Fig. 1. Hardware of OOK shear wave acoustic communication on
pipesetup. (1) Digital computer with GNURadio software. (2) Red
Pitayatransmitter board. (3) Power amplifier. (4) Angled-wedge
mounted PZTtransmitting refracted shear waves. (5) Stainless steel
pipe. (6) Angled-wedge mounted PZT receiving shear waves. (7)
Low-noise amplifier.(8) Red Pitaya receiver board. (9) Digital
oscilloscope.
ultrasonic transducers, such as high-temperature
compatibleelectromagnetic transducers (EMATs), can be used for
datatransmission, PZT has better coupling efficiency for
non-ferromagnetic pipes. The PZT is mounted on a commercialacrylic
angled wedge, and the angle of which exceeds the firstcritical
angle of 27.6◦ for the acrylic/stainless steel interface.Although
the nominal frequency of the PZTs is 2 MHz, itwas determined that,
once coupled to the pipe through angledwedges, the largest
amplitude signal was obtained by operatingPZT at 1.8 MHz.
Communication with shear waves is advantageous becausethey do
not couple into the water, which could be presentinside the pipe.
In addition, the excitation of multiple modesthat travel at
different velocities along the pipe can leadto intersymbol
interference. Below the first critical angle,a refracted wave
consists of both longitudinal and shear wavecomponents, while above
the critical angle only shear waveis refracted. The study utilized
a commercial 45◦ wedge,for which it was determined experimentally
that the receivedsignal consisted of shear waves only.
In principle, in nondestructive testing applications,
elasticwaves on pipes are frequently generated with the
radiallysymmetric collar-type transducer. However, the CVCS
pipesare part of the thermal-hydraulic system, and as such
areenclosed by a layer of thermal insulation. Thermal
insulationmaterials, such as mineral wool, are poor transmitters
ofacoustic waves. To achieve efficient coupling of acoustic
wavesinto the pipe, the insulation material has to be removed for
thetransducer to be directly in contact with the metal pipe.
Thus,it is preferable to use a transducer with the smallest
formfactor so that minimal amount of thermal insulation needs tobe
removed and no thermal imbalances in the coolant systemare
created.
A schematic depiction of the communication system isshown in
Fig. 1. A digital signal is generated by theGNURadio program (1),
which is next converted into ananalog signal through modulation of
the amplitude of the
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
1194 IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND
FREQUENCY CONTROL, VOL. 67, NO. 6, JUNE 2020
carrier ultrasonic wave by Red Pitaya electronic board (2).In
our study, we encoded information using binary ASK. Theanalog wave
is amplified with a 50-dB power amplifier (3)and converted into an
ultrasonic pressure wave with a PZT(4). The incident longitudinal
wave is mode-converted into ashear wave, which subsequently
propagates down the stainlesssteel pipe (5). By symmetry, the shear
wave is refracted intothe angled wedge at the receiving end of the
pipe to becomea longitudinal wave, which is converted by the
receiving PZT(6) into an electrical signal. The received signal is
amplifiedwith a 20-dB low noise amplifier (LNA) and demodulated
withreceiver Red Pitaya board (8) and passed through a
low-passfilter to recover the information encoded in the amplitude
ofthe carrier. The analog signal is decimated to create a
digitalsignal, and the bits are repacked with GNURadio software.In
this study, transmitted signals from PA and received signalsfrom
LNA were sampled with a digital oscilloscope (9) toanalyze the
performance of the communication protocol fortransmission of
images.
Preliminary studies have shown that ultrasonic informa-tion
transmission over the pipe is resilient to low-frequencynoise.
Proof-of-principle demonstrations were performed withmechanical
shaker vibrating the pipe at frequencies from100 Hz to 10 KHz, with
no observable effects on informationtransmitted with 1.8-MHz
carrier shear wave [5].
B. Computer Modeling of RefractedShear Wave Coupling
Computer simulations were performed with the COMSOLMultiphysics
Solid Mechanics Module software package tomodel generation of
refracted shear waves with an angledwedge-mounted transducer on a
CVCS-like pipe.
In the COMSOL model, refracted shear waves are generatedat the
boundary of the acrylic wedge and stainless steelpipe through a
direct solution of elastodynamic equations inCOMSOL. Instead of
explicitly modeling PZT, the source oflongitudinal waves is a
boundary load applied to the acrylicwedge surface, which would in
contact with the PZT surfacein the experiment. As could be seen in
the graphics in Fig. 2,PZT structure is absent from the COMSOL
model. Ultrasonicfrequency of 500 KHz was used in the computer
simulationsto reduce memory requirements due to coarser grid
meshingfor longer wavelength. While this frequency is smaller than
the2 MHz of the transducer used in the experiment, the physicsof
refracted shear wave generation and transmission does notchange
appreciably with frequency. The pipe in the COMSOLmodel has the
same diameter and thickness as the experimentalarticle (2.375 in
outer diameter and 0.344 in wall thickness),but the length is
shortened to 60 cm (2 ft). The amplitudeof the ultrasonic elastic
wave is visualized with a pseudo-color map of pressure
distribution, with green being zero,and warmer and colder colors
corresponding to positive andnegative values, respectively. To
increase the visibility ofwavefronts in this coloring scheme, the
amplitude of the elasticwave was taken to be 4 × 107 N/m2.
Fig. 2 (top) shows the pressure distribution of 40 µs afterthe
start of the simulations when the refracted shear waveis coupled
into the pipe. Fig. 2 (middle) shows the pressure
Fig. 2. Computer simulations of coupling and propagation of
refractedshear wave on the pipe, excited with 500-kHz PZT on a 45◦
angle wedge.Ultrasonic wavefronts are visualized with a
pseudo-color map of pressuredistribution at 40, 110, and 200
µs.
distribution after 110 µs when the wave reaches approximatelythe
middle of the pipe (30 cm mark). Since the shear wavevelocity in
stainless steel is 3100 m/s, after 110 µs the prop-agation distance
is 34 cm, which is qualitatively in agreementwith COMSOL
simulations. The longitudinal wave velocityis 5790 m/s so that in
110 µs the longitudinal wave wouldhave traversed 63 cm distance,
which is the entire lengthof the pipe. This is not observed in
COMSOL simulationsshown in Fig. 2. Therefore, these observations
confirm thatthe COMSOL model generates refracted shear waves on
apipe. Fig. 2 (bottom) shows the pressure distribution at 200
µsafter the start of the simulations. The pseudo-color map
ofpressure distribution indicates that the wave has reached theend
of the pipe. This is consistent with estimations basedon shear wave
velocity, which predict propagation distanceof 62 cm after 200 µs.
This simulation confirms qualitativelythat pure ultrasonic shear
wave is generated in the experimen-tal configuration.
Visualizations of ultrasonic wavefronts in Fig. 2 indicatethat
while the wave is initially propagating at an angle to the
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
HEIFETZ et al.: TRANSMISSION OF IMAGES WITH ULTRASONIC ELASTIC
SHEAR WAVES 1195
pipe axis, after propagating for approximately 30 cm
distance,the wave is collinear with the pipe axis. For example, in
thefar field, the wave from a single transducer has a similar
radialprofile as that generated by a ring of transducers.
Experimentson transmitting single pulses across the pipe have
shownthat the amplitude of the received signal does not
changeappreciably when the receiving transducer is positioned at
90◦and 180◦ relative to the transmitting PZT. This indicates
thatthe communication system is resilient to the misalignment
oftransducers.
C. Overview of GNURadio Communication ProtocolThe data
transmission program was implemented in
GNURadio, which is a freeware Defined Radio (SDR) pro-gramming
environment. The flowchart of the ASK communi-cation program
developed in GNURadio is shown in Fig. 3.The program consists of
blocks performing modulation anddemodulation functions. While more
elaborate forms of shiftkeying exist that allow for more
information to be encodedper key, we chose ASK technique due to the
fact that theconstellation symbols are maximally spread out,
reducing theimpact of noise on the channel compared to a setup
thatencodes more than one bit per symbol.
ASK communication consists of a binary stream of infor-mation.
Since an image data has a non-binary format, we havechosen a
portable pixel map (PPM) file type to be the imagedata structure.
The PPM file type stores the image dimensionsin the first three
lines. The rest of the PPM file contains theimage stored as a
matrix of ASCII character. We will refer tothe three lines of the
PPM file as the “image header” and callthe rest as the “image
payload.” Error-free transmission of theimage header is critical to
the successful reconstruction of thereceived image. The imaging
payload, on the other hand, hasa lower tolerance to errors since a
few errors are not likely tomake the file unreadable.
Transmission of the image involves the conversion of thePPM,
data structure into a binary one, and reassembling thereceived
binary data into the original PPM format. To accom-plish these
tasks, we first disassemble the file into a stream ofASCII
characters, in the form of bytes (1 byte = 8 bits). Thesecharacters
are then converted into a bitstream with an attachedpacket header.
After demodulation, the bit stream is parsed toremove the packet
header, the bits are converted into theiroriginal bytes, and the
file is reassembled. If error correctionis added to the protocol,
after the bytes are converted intobits, they are multiplied by a
convolution matrix. Once thecommunication packet is received, the
bits are decoded by aninverse matrix.
III. ANALYSIS OF GNURADIO COMMUNICATIONPROTOCOL PERFORMANCE
A. Communication Protocol Testing and AnalysisTo analyze the
performance of the data transmission pro-
gram, a simple message with characters “!s!” was passedthrough
the system and monitored in the output of each blockin the
flowchart in Fig. 3. Output signals from select blocks inthe
communication program, plotted with Python Matplotlibsoftware, are
displayed in Figs. 4–11.
Fig. 3. ASK communication protocol signal flow graph created
inGNURadio. Each block is numbered at the output.
In the File Source block #1, the file is disassembled intothree
ASCII values: 33, 115, 33. The value of 33 corre-sponds to “!” and
the value of 115 corresponds to “s.” TheProtocol Formatter block #4
creates a tagged stream of 1’sand 0’s that will be attached to the
payload. We use thedefault preamble of “1010010011110010” and
access codeof
“1010110011011101101001001110001011110010100011000010000011111100.”
Fig. 4 visualizes the result of combin-ing the header file with the
payload in Tagged Stream MUXblock #5. The first 12 bytes (0–11) are
from the protocolformatter, while the last three (12, 13, and 14)
correspond tothe ASCII values of the characters “!s!” from the file
source.
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
1196 IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND
FREQUENCY CONTROL, VOL. 67, NO. 6, JUNE 2020
Fig. 4. Tagged Stream MUX block #5 output (data colored as
individualbytes).
Fig. 5. Analog square wave at the Moving Average block #10
output(data colored as individual bytes).
The Repack Bits block #6 unpacks each byte into eightbits. There
are 15 bytes in the packet, which means that120 samples are sent
through the system (15∗8 = 120).Fig. 5 shows the signals at the
output of the Moving Averageblocks #10, following the Rational
Resampler block #9. Thedigital signal generated with Repack Bits
block #6 is convertedinto an analog square wave. Each bit is
interpolated to create600 bits so that the total number of samples
increases from12 to 72 000 (120∗600 = 72 000).
The spectrum of the time-domain signal in Fig. 5, calculatedvia
digital Fast Fourier Transform (FFT), indicates that theeffective
bandwidth of the square wave is less than 50 kHz.Fig. 6 displays
the transmitted wave output of Red Pitaya Sinkblock #12. The
carrier frequency is 1.8 MHz. The data inFig. 6 is recorded with
the digital oscilloscope (#9 in Fig. 1)at a sampling rate of 25
Ms/s, which is much higher thanGNURadio’s rate of 100 Ks/s. The
number of samples forthe same signals is 250 times higher than in
GNURadio(250 ∗ 100 000 = 25 000 000). The individual bytes from
thesquare wave are all translated into a signal with amplitude
Fig. 6. Transmitted wave output of Red Pitaya Sink block #12,
recordedwith a digital oscilloscope (data colored as individual
bytes).
Fig. 7. Received wave on the pipe before demodulation in Red
PitayaSource block #13 (data colored as individual bytes).
ranging from −0.3 to 0.3 V. Note that in the experimentalsetup
in Fig. 1, this signal passes through a 50-dB poweramplifier (#3),
and the PZT (#4) senses a signal with approx-imately 95-V
amplitude. One can observe in Fig. 6 that theRed Pitaya board
introduces ringing noise into the system inevery pulse ON/OFF
transition. This behavior is the reason whybinary ASK is preferable
to quad ASK since ringing noisewould be amplified in each
additional OOK event.
Fig. 7 displays the received signal waveform from RedPitaya
Source block #13, but before demodulation is per-formed. The data
are recorded with the digital oscilloscope(#9 in Fig. 1). The
received signal is amplified with an LNA(#7 in Fig. 1) but still
appears to be attenuated by an orderof magnitude relative to the
transmitted signal in Fig. 6.In addition, a dc offset of −0.075 V
is added to the signal.Compared to the transmitted signal in Fig.
6, there are moredistortions on the peaks of the received waveform,
as well asa much larger amount of noise around the dc offset.
Fig. 8 shows the received signal with the Red Pitayaboard
demodulated with Red Pitaya Source block #13.
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
HEIFETZ et al.: TRANSMISSION OF IMAGES WITH ULTRASONIC ELASTIC
SHEAR WAVES 1197
Fig. 8. Received signal with Red Pitaya board after demodulation
withRed Pitaya Source block #13 (data colored as individual
bytes).
The system produces a noisy waveform lacking informa-tion
(sample numbers
-
1198 IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND
FREQUENCY CONTROL, VOL. 67, NO. 6, JUNE 2020
Fig. 11. GNURadio screen capture of recovered received image
ofANL logo.
propagation over complex piping structures. Detailed analysisof
practical deployment considerations is beyond the scope ofthe
present study. In this section, we briefly review severalpossible
challenges and their potential resolution.
In principle, piping deterioration, in particular, corrosionand
cracking, could have a negative impact on the data trans-mission
performance. However, the condition of the nuclearfacility pipes is
rigorously maintained through controlling pHand filtering out
debris in the fluid, as well as performingthorough inspections
during regular reactor shutdown periods.Therefore, deterioration of
piping is not expected to have amajor impact on the ultrasonic
communication at a nuclearfacility.
The study in this article focused on ultrasonic data
trans-mission on a straight pipe. In principle, a straight section
ofpiping might not be available for mounting transducers.
Pipingelbows and bends are fairly common at nuclear facilities,
andtransmission of over such piping manifolds could be necessaryto
connect specific locations in the facility with a piping
com-munication channel. We conducted a preliminary evaluation ofthe
signal transmission across a bent piping structure with a90◦ turn
using COMSOL computer simulations. The model ofthe structure,
consisting of two 30-cm-long straight sectionsjoined with an elbow,
is shown in Fig. 12. All other parametersof the metallic structure
in the COMSOL model are the sameas those in Fig. 2. Longitudinal
ultrasonic waves at 500-kHzfrequency were coupled as boundary
pressure load to theacrylic wedge angled at 45◦. Elastic waves are
visualized witha pseudo-color plot of pressure distribution. The
amplitudeof the pressure is amplified compared to actual
experimentalvalues to enhance the elastic wave visibility.
Fig. 12 (top) shows the pressure distribution of 40 µs afterthe
start of the simulations when the refracted shear waveis coupled
into the pipe. Fig. 12 (middle) shows pressuredistribution after
110 µs when the wave reaches the elbow.Note that the propagation
distance to the geometrical centerof the bent section is slightly
larger than 30 cm because theelbow increases the length of the
overall piping structure byapproximately 10 cm. Fig. 12 (bottom)
shows the pressuredistribution at 200 µs after the start of the
simulations.At this point, the ultrasonic shear wave has propagated
acrossthe elbow and reached the middle of the second
straightsection.
Fig. 12. Propagation of 500-kHz refracted shear wave on a
metallicbent pipe, visualized with pseudo-color map of pressure
distribution at40, 110, and 200 µs.
Since the shear wave velocity in stainless steel is 3100 m/s,in
110 and 200 µs the propagation distances are 34 and 62
cm,respectively, which is in qualitative agreement with
COMSOLsimulations. No reverberations or scattering from the
elbowbend are observed in computer simulations. This
confirmsqualitatively that ultrasonic shear waves travel across
uniformpiping bend without mode conversion, which is consistent
withprior studies [22].
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
HEIFETZ et al.: TRANSMISSION OF IMAGES WITH ULTRASONIC ELASTIC
SHEAR WAVES 1199
However, it should be noted that piping bends are
typicallyformed by welding of several piping sections. The
weldsintroduce discontinuities in the piping material, which
cancause significant scattering of the ultrasonic shear wave. A
testarticle consisting of a stainless steel pipe bent at 90◦
wasdeveloped for preliminary laboratory analysis. The bent
pipingtest article developed by welding two straight
three-foot-long pipes to an elbow. For consistency, the diameter
andwall thickness of the bent pipe is the same as that of
thestraight pipe (6-ft-long schedule 160 pipes with 2.375-in
outerdiameter). Industrial-grade welding was performed to
achieveleak proof under the 2000 psi pressure test. After welding,
theouter surface was ground to achieve a visibly smooth
finish.Preliminary results of 500-kHz ultrasonic refracted shear
wavetransmission across the bent piping structure indicate thatthe
shear wave is significantly distorted by the welds. Onepossible
mitigation strategy is to use time-reversal modulation(TRM) to
remove noise from the received signal. Integration ofTRM filter
into ultrasonic communication on complex pipingmanifolds will be
investigated in future studies.
V. CONCLUSION
In this article, we discussed the design and
performanceevaluation of an ultrasonic communication system on
anuclear-grade stainless steel pipe. A laboratory bench-scalesystem
consisting of a nuclear-grade chemical volume controlsystem
(CVCS)-like pipe and commercial PZT ultrasonictransducers were
assembled for a preliminary communicationsystem demonstration.
Carriers of information are ultrasonicrefracted shear waves on the
pipe. ASK communicationprotocol for image transmission was
developed using theGNURadio software environment. Detailed analysis
of thecommunication protocol, including the output of each block,is
presented in this article. Using the communication system,a 32-KB
image was transmitted at a data bitrate of 2 Kbpsacross the
stainless steel pipe. The next phase of this workwill investigate
the strategy for increasing communicationbitrate. In addition,
instead of PZTs designed for operation,data transmission with
high-temperature ultrasonic transducerswill be investigated.
Finally, signal processing strategies forultrasonic signal
transmission over complex piping manifoldswill be investigated.
REFERENCES
[1] A. Heifetz, R. B. Vilim, and S. Bakhtiari, “Transmission of
informationby acoustic communication along metal pathways in
nuclear facilities,”U.S. Patent 15 947 303, Sep. 13, 2018.
[2] W. F. Young, C. L. Holloway, G. Koepke, D. Camell, Y.
Becquet, andK. A. Remley, “Radio-wave propagation into large
building structures—Part 1: CW signal attenuation and variability,”
IEEE Trans. AntennasPropag., vol. 58, no. 4, pp. 1279–1289, Apr.
2010.
[3] G. Masche, Systems Summary of Westinghouse Pressurized Water
Reac-tor Nuclear Power Plant. Pittsburgh, PA, USA: Westinghouse
ElectricCorp., 1971.
[4] E. Leinov, M. J. Lowe, and P. Cawley, “Investigation of
guided wavepropagation and attenuation in pipe buried in sand,” J.
Sound Vib.,vol. 347, pp. 96–114, Jul. 2015.
[5] A. Heifetz, X. Huang, R. Ponciroli, S. Bakhtiari, R. Vilim,
and J. Saniie,“Transmission of information using elastic waves on
metal pipes innuclear facilities,” in Proc. Nucl. Plant Instrum.,
Control Human-Mach.Interface Technol. (NPIC&HMIT), 2019, pp.
49–58.
[6] M. Stojanovic and J. Preisig, “Underwater acoustic
communicationchannels: Propagation models and statistical
characterization,” IEEECommun. Mag., vol. 47, no. 1, pp. 84–89,
Jan. 2009.
[7] A. Sing, J. Nelson, and S. Kozat, “Signal processing for
underwa-ter acoustic communications,” IEEE Commun. Mag., vol. 47,
no. 1,pp. 90–96, Jan. 2009.
[8] C. Li, D. Hutchins, and R. Green, “Short-range ultrasonic
digital com-munications in air,” IEEE Trans. Ultrason.,
Ferroelectr., Freq. Control,vol. 55, no. 4, pp. 908–918, Apr.
2008.
[9] C. Li, D. Hutchins, and R. Green, “Short-range ultrasonic
communi-cations in air using quadrature modulation,” IEEE Trans.
Ultrason.,Ferroelectr., Freq. Control, vol. 56, no. 10, pp.
2060–2072, Oct. 2009.
[10] J. D. Ashdown, L. Liu, G. J. Saulnier, and K. R. Wilt,
“High-rateultrasonic through-wall communications using MIMO-OFDM,”
IEEETrans. Commun., vol. 66, no. 8, pp. 3381–3393, Aug. 2018.
[11] B. Wang, J. Saniie, S. Bakhtiari, and A. Heifetz,
“Architectureof an ultrasonic experimental platform for information
transmissionthrough solids,” in Proc. IEEE Int. Ultrason. Symp.
(IUS), Sep. 2017,pp. 1–4.
[12] M. Malzer, C. Kexel, T. Maetz, and J. Moll, “Combined
inspection anddata communication network for Lamb-wave structural
health monitor-ing,” IEEE Trans. Ultrason., Ferroelectr., Freq.
Control, vol. 66, no. 10,pp. 1625–1633, Oct. 2019.
[13] J. L. Rose, Ultrasonic Guided Waves in Solid Media.
Cambridge, U.K.:Cambridge Univ. Press, 2014.
[14] S. Chakraborty, G. J. Saulnier, K. W. Wilt, E. Curt, H. A.
Scarton, andR. B. Litman, “Low-power, low-rate ultrasonic
communications systemtransmitting axially along a cylindrical pipe
using transverse waves,”IEEE Trans. Ultrason., Ferroelectr., Freq.
Control, vol. 62, no. 10,pp. 1788–1796, Oct. 2015.
[15] Y. Jin, Y. Ying, and D. Zhao, “Data communications using
guided elasticwaves by time reversal pulse position modulation:
Experimental study,”Sensors, vol. 13, no. 7, pp. 8352–8376, Jul.
2013.
[16] H. Sakuma, K. Nakamura, and S. Ueha, “Two-way communication
overgas pipe-line using multicarrier modulated sound waves with
cyclicfrequency shifting,” Acoust. Sci. Tech., vol. 27, no. 4, pp.
225–232, 2006.
[17] X. Huang, J. Saniie, S. Bakhtiari, and A. Heifetz,
“Ultrasonic commu-nication system design using electromagnetic
acoustic transducer,” inProc. IEEE Int. Ultrason. Symp. (IUS), Oct.
2018, pp. 1–4.
[18] X. Huang, J. Saniie, S. Bakhtiari, and A. Heifetz,
“Applying EMAT forultrasonic communication through steel plates and
pipes,” in Proc. IEEEInt. Conf. Electro/Inf. Technol. (EIT), May
2018, pp. 379–383.
[19] B. Wang, J. Saniie, S. Bakhtiari, and A. Heifetz, “Software
definedultrasonic system for communication through solid
structures,” in Proc.IEEE Int. Conf. Electro/Inf. Technol. (EIT),
May 2018, pp. 267–270.
[20] B. Wang, J. Saniie, S. Bakhtiari, and A. Heifetz, “A
high-performancecommunication platform for ultrasonic
applications,” in Proc. IEEE Int.Ultrason. Symp. (IUS), Oct. 2018,
pp. 1–4.
[21] B. Wang, J. Saniie, S. Bakhtiari, and A. Heifetz,
“Ultrasonic communi-cation systems for data transmission,” in Proc.
IEEE Int. Conf. ElectroInf. Technol. (EIT), May 2019, pp.
479–482.
[22] P. Petcher, S. Burrows, and S. Dixon, “Shear horizontal
(SH) ultrasoundwave propagation around smooth corners,”
Ultrasonics, vol. 54, no. 4,pp. 997–1004, Apr. 2014.
Alexander Heifetz received the B.S. degree(summa cum laude) in
applied math, the M.S.degree in physics, and the Ph.D. degree in
elec-trical engineering from Northwestern University,Evanston, IL,
USA, in 1999, 2002, and 2006,respectively.
He joined the Nuclear Engineering Division,Argonne National
Laboratory, Argonne, IL, USA,as Director’s Postdoctoral Fellow, in
2008, andbecame an Electrical Engineer (Technical StaffMember) in
2011, and the Principal Electrical
Engineer with Nuclear Science and Engineering Division in
2017.At Argonne, he has been involved in projects related to
nuclearenergy enabling technology, including ultrasonic
communications, ther-mal tomography nondestructive evaluation for
additive manufacturing,and thermal hydraulics sensing and
control.
Dr. Heifetz was a co-recipient of the Best Paper Award at 2019
IEEEInternational Conference on Electro-Information Technology
(EIT) andhas given invited talks at the Purdue School of Nuclear
Engineeringin 2019 and at the UIC Department of Civil and Materials
Engineeringin 2017.
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
-
1200 IEEE TRANSACTIONS ON ULTRASONICS, FERROELECTRICS, AND
FREQUENCY CONTROL, VOL. 67, NO. 6, JUNE 2020
Dmitry Shribak received the B.S. degree inphysics and molecular
engineering from the Uni-versity of Chicago, Chicago, IL, USA, in
2019.
He was a Science Undergraduate LaboratoryIntern (SULI) with
Nuclear Science and Engi-neering Division, Argonne National
Laboratory,Argonne, IL USA. He is currently a GraduateStudent with
the Department of Electrical Engi-neering, Georgia Institute of
Technology, Atlanta,GA, USA. His current research interests
includeRF wave propagation, simulation, and antennadesign.
Xin Huang received the B.S. degree in com-munication engineering
from Zhengzhou Univer-sity, Zhengzhou, Henan, China, in 2012,
andthe M.S. degree in sensor technology from theUniversity of
Shanghai for Science and Technol-ogy, Shanghai, China, and the M.E.
degree insensor technology from the Coburg Universityof Applied
Science, Coburg, Germany, in 2015.He is currently pursuing the
Ph.D. degreewith Embedded Computing and Signal Process-ing Research
Laboratory, Illinois Institute of
Technology, Chicago, IL, USA.He is currently a Research
Assistant in the Electrical and Computer
Engineering Department, Illinois Institute of Technology. He was
theGraduate Student Research Assistant in the Nuclear Science and
Engi-neering Division, Argonne National Laboratory, Argonne, IL
USA, from2017 to 2019. His current research interests include
ultrasonic sensors,ultrasonic communications, software-defined
radio, modeling and signalanalysis, machine learning, embedded
computing, and system-on-chipdesign.
Boyang Wang received the B.S. degree ininformation engineering
from the Beijing Instituteof Technology, Beijing, China, in 2013,
and theM.S. degree in electrical engineering from theIllinois
Institute of Technology, Chicago, IL USA,in 2015, where he is
currently pursuing the Ph.D.degree.
He is currently a Research Assistant in theElectrical and
Computer Engineering Depart-ment, Illinois Institute of Technology.
His researchinterests include ultrasonic signal processing,
software-defined radio, machine vision, automation, internet of
things,embedded system development, hardware, and software
co-design,system-on-chip design, and artificial intelligence and
machine learning.
Jafar Saniie (Life Fellow, IEEE) received the B.S.degree (Hons.)
in electrical engineering from theUniversity of Maryland, College
Park, MD, USA,in 1974, the M.S. degree in biomedical engi-neering
from Case Western Reserve University,Cleveland, OH, USA, in 1977,
and the Ph.D.degree in electrical engineering from
PurdueUniversity, West Lafayette, IN, USA, in 1981.
In 1981, he joined the Department of AppliedPhysics, University
of Helsinki, Finland, to con-duct research in photothermal and
photoacoustic
imaging. Since 1983, he has been with the Department of
Electrical andComputer Engineering, Illinois Institute of
Technology, Chicago, IL USA,where he is the Department Chair, the
Filmer Endowed Chair Profes-sor, and the Director of Embedded
Computing and Signal Processing(ECASP) Research Laboratory. He has
over 340 publications and hassupervised 35 Ph.D. dissertations, and
22 M.S. theses to completion.His research interests and activities
are in ultrasonic signal and imageprocessing, ultrasonic
software-defined communication, artificial intelli-gence and
machine learning, statistical pattern recognition, estimationand
detection, data compression, time–frequency analysis,
embeddeddigital systems, system-on-chip hardware/software
co-design, Internetof things, computer vision, deep learning, and
ultrasonic nondestructivetesting and imaging.
Dr. Saniie has been a Technical Program Committee Member of
theIEEE Ultrasonics Symposium since 1987 (The Chair of Sensors,
NDEand Industrial Applications Group, 2004–2013), has been an
AssociateEditor of the IEEE TRANSACTIONS ON ULTRASONICS,
FERROELECTRICS,AND FREQUENCY CONTROL since 1994, the Lead Guest
Editor for theIEEE ULTRASONICS, FERROELECTRICS AND FREQUENCY
CONTROL (UFFC)Special Issue on Ultrasonics and Ferroelectrics
(August 2014), the IEEE
UFFC Special Issue on Novel Embedded Systems for Ultrasonic
Imagingand Signal Processing (July 2012), and Special Issue on
Advancesin Acoustic Sensing, Imaging, and Signal Processing
published in theJournal of Advances in Acoustics and Vibration,
2013. He received the2007 University (Illinois Institute of
Technolgy) Excellence in TeachingAward. He was the General Chair of
the 2014 IEEE Ultrasonics Sympo-sium in Chicago. He has served as
the IEEE UFFC Ultrasonics AwardsChair since 2018. He served as the
Ultrasonics Vice President for theIEEE UFFC Society
(2014–2017).
Jacey Young received the B.S. degree inphysics from St. Norbert
College, De Pere, WI,USA, in 2019.
She interned at Brigham Young University,Provo, UT, USA, where
she worked on usingsound pressure to locate acoustical sources.As a
Science Undergraduate Laboratory Intern(SULI) at Argonne National
Laboratory, Lemont,IL, USA, she was involved in the
theoreticalcalculations for acoustical transmission throughsteel
pipes for the United States Department of
Energy. In 2019, she joined Epic Systems Corporation, Verona,
WI, USA,as a Software Developer.
Sasan Bakhtiari (Senior Member, IEEE)received the B.S.E.E.
degree from the IllinoisInstitute of Technology, Chicago, IL USA,in
1983, the M.S.E.E. degree from the Universityof Kansas, Lawrence,
KS, USA, in 1987, andthe Ph.D. degree in electrical engineering
fromColorado State University, Fort Collins, CO,USA, in 1992.
Since 1993, he has been with ArgonneNational Laboratory, Lemont,
IL, USA, where heis currently a Senior Electrical Engineer and
the
Section Manager of the Sensors, Instrumentation and NDE Group
inthe Nuclear Science and Engineering Division. He has been
involvedwith applied research in the areas of electromagnetic
guided-wave andradiating structures, induction sensor technology,
acoustic sensing,signal processing, and analytical and numerical
modeling. He hasconducted theoretical and experimental work on
electromagnetic andacoustic/ultrasonic NDE techniques as well as
active and passivemillimeter-wave sensing techniques for various
industrial and scientificapplications. He has authored more than
150 journal articles, conferenceproceedings, technical reports, and
chapter contributions.
Dr. Bakhtiari was a recipient of two Research and Development100
Awards and the Inventor of a number of patents related tononcontact
electromagnetic and acoustic sensing technologies.
Richard B. Vilim received the B.Eng. degreein engineering
physics from McMaster Univer-sity, Hamilton, ON, Canada, in 1976,
the M.Sc.degree in electrical engineering from the Uni-versity of
Toronto, Toronto, ON, in 1978, andthe Ph.D. degree in nuclear
engineering fromthe Massachusetts Institute of Technology
(MIT),Cambridge, MA, USA, in 1983.
He is currently a Senior Nuclear Engineer andmanages the Plant
Analysis and Control andNondestructive Evaluation Sensors Group
within
Argonne’s Nuclear Science and Engineering Division. He is also
anExpert on modeling and simulation of power plants from both an
engi-neering and economic perspective, on power plant
instrumentation andcontrol, and on equipment monitoring, with over
35 years of experiencedeveloping new applications for the process
and power industries. He ispresently leading the team responsible
for designing the instrumenta-tion and control system for an
advanced nuclear power plant to bebuilt with international
collaboration. His research focuses on the costcompetitiveness of
nuclear power operating in both an electric grid withrenewables and
in hybrid energy systems. His research also includes thedevelopment
of signal-processing methods and automated reasoningmethods for
detecting sensor failure and for the diagnosis of
incipientequipment degradation in process systems. He has over 250
publica-tions. He holds seven U.S. patents. He has developed a
number ofcomputer codes in use by others including General Plant
Analyzer andSystem Simulator (GPASS) and Parameter-Free Reasoning
Operatorfor Automated Identification and Diagnosis (PRO-AID). He is
one offour Argonne inventors whose patents formed the intellectual
propertybasis for Smart Signal Corporation, a technology startup
company withrevenues of $30M, recently acquired by General
Electric.
Authorized licensed use limited to: Illinois Institute of
Technology. Downloaded on September 30,2020 at 21:12:06 UTC from
IEEE Xplore. Restrictions apply.
/ColorImageDict > /JPEG2000ColorACSImageDict >
/JPEG2000ColorImageDict > /AntiAliasGrayImages false
/CropGrayImages true /GrayImageMinResolution 150
/GrayImageMinResolutionPolicy /OK /DownsampleGrayImages true
/GrayImageDownsampleType /Bicubic /GrayImageResolution 600
/GrayImageDepth -1 /GrayImageMinDownsampleDepth 2
/GrayImageDownsampleThreshold 1.50000 /EncodeGrayImages true
/GrayImageFilter /DCTEncode /AutoFilterGrayImages true
/GrayImageAutoFilterStrategy /JPEG /GrayACSImageDict >
/GrayImageDict > /JPEG2000GrayACSImageDict >
/JPEG2000GrayImageDict > /AntiAliasMonoImages false
/CropMonoImages true /MonoImageMinResolution 300
/MonoImageMinResolutionPolicy /OK /DownsampleMonoImages true
/MonoImageDownsampleType /Bicubic /MonoImageResolution 900
/MonoImageDepth -1 /MonoImageDownsampleThreshold 1.33333
/EncodeMonoImages true /MonoImageFilter /CCITTFaxEncode
/MonoImageDict > /AllowPSXObjects false /CheckCompliance [ /None
] /PDFX1aCheck false /PDFX3Check false /PDFXCompliantPDFOnly false
/PDFXNoTrimBoxError true /PDFXTrimBoxToMediaBoxOffset [ 0.00000
0.00000 0.00000 0.00000 ] /PDFXSetBleedBoxToMediaBox true
/PDFXBleedBoxToTrimBoxOffset [ 0.00000 0.00000 0.00000 0.00000 ]
/PDFXOutputIntentProfile (None) /PDFXOutputConditionIdentifier ()
/PDFXOutputCondition () /PDFXRegistryName () /PDFXTrapped
/Unknown
/CreateJDFFile false /Description >>>
setdistillerparams> setpagedevice