As fossil fuel depletion and environmental pollution problems are becoming increasingly more serious, interest in the efficient use of natural resources and alternative energy is rapidly growing. In particular, interest in fuels stored as high-pressure gases such as natural gas and hydrogen is also rising. Ultrasonic waves show various received signals according to characteristics such as density of the medium and acoustic impedance. An experimental study on the detection of the micro-leakage of fuel stored as high-pressure gas was conducted based on the characteristics of ultrasonic waves. First, an ultrasonic sensor was manufactured by selecting the matching layer with consideration of the acoustic impedance. In the experiment, a mass flow controller (MFC) was attached to a perforated hole in the fabricated chamber to generate micro-leakage, and the signal from the receiving ultrasonic sensor was then collected. The envelope signal of the received ultrasonic sensor signal was analyzed through the Gaussian distribution method. The temperature inside the chamber and the received voltage decreased according to a similar trend and showed a nonlinear result. However, the phase of the received ultrasonic sensor signal showed a relatively linear result according to the internal pressure change. Micro-leakage could not be detected with only the received voltage seen by the ultrasonic sensor. Therefore, the phase shift of the receiving ultrasonic sensor can be used to detect micro-leakage in a high-pressure gas tank.
In recent years, problems such as environmental pollution and the exhaustion
of petroleum resources have been emerging around the world. As a part of
efforts to address these problems, studies have been conducted on thermal
efficiency and the emission characteristics of hydrogen/compressed natural
gas mixture (HCNG) spark-ignition engines (Ma et al., 2007), and comparative
studies have been conducted on compressed natural gas with direct injection
(CNG-DI), compressed natural gas with bi-fuel direct injection (CNG-BI), and
gasoline with port injection (gasoline-PI) engines (Kalam and Masjuki, 2011). As
evidenced by the aforementioned studies, interest in and research on natural
gas and alternative energy are increasing rapidly. Among natural resources,
natural gas has the same thermal efficiency as fossil fuels and causes less
pollution (Hesterberg et al., 2008). In addition, among alternative
energies, hydrogen does not emit pollutants such as carbon dioxide, and
there is specific interest in fuels stored in gaseous form (Jacobson et al.,
2005). Natural gas has a high calorific value and has the advantage of low
emissions of particulate matter (PM), nitrogen oxides (
Ultrasound has various reception characteristics depending on the density and acoustic impedance of the measurement medium, and, on the basis of these characteristics, it is widely used in communication devices, medical devices, actuators, transducers, precision sensors, nondestructive testing (NDT), and measuring devices (Vellekoop, 1998; Choi et al., 2020). In addition, it has been used in structural health monitoring (SHM) systems (Guo et al., 2020), liquid level monitoring systems (Gao et al., 2020), early crack detection of structural systems (Chakraborty et al., 2019), and gas leak location detection systems (Wang et al., 2018). In this paper, an experimental study was conducted to analyze the characteristics of ultrasonic waves for the measurement of micro-leakage in a high-pressure gas tank instead of the conventional pressure sensor.
Waves are generated by elasticity against compression deformation of the
particles constituting the medium through physical vibration, and this is
called a sound wave. Sound waves can be classified into longitudinal waves,
transverse waves, coda waves (vibration that persists for a long time even after the arrival time of the surface wave), and surface waves according to the direction
of vibration. In addition, they can be classified into three ranges
according to the frequency. Vibration frequencies below 20 Hz are defined as
low frequencies, vibration frequencies in the 20 Hz–20 kHz
range are defined as audible frequencies that humans can hear, and
frequencies above 20 kHz are defined as ultrasonic (Cheeke, 2017). Ultrasonic waves carry wave energy, and phenomena such as dispersion and diffraction occur. In addition,
in the process of being transferred to another medium, some energy may be
lost and attenuation may occur. Ultrasonic waves move similarly to light
waves, but, unlike light waves that can move in a vacuum, an elastic medium
is required (Sharma et al., 2017). In order to artificially generate
ultrasonic waves, a sound-emitting body with a high frequency is required.
The sound-emitting body uses the electrical properties of an object to apply
a high-frequency alternating voltage to both ends of the object to generate
ultrasonic waves through the vibration of the object. Currently, the most
commonly used sound-emitting body is lead zirconate titanate made by mixing
lead titanate (
The piezoelectric element and the matching layer are the key factors of
ultrasonic sensors. Alvarez-Arenas (2004) investigated
the requirements to select the material of the piezoelectric element
laminated layer. It was found that the selection of materials with high
porosity and fine pores and the thickness of the matching layer affect the
performance of the ultrasonic sensor. Kim et al. (2019) and
Kim and Lee (2019) conducted a study on the envelope signal change of the
received signal according to the composition of the matching layer and the
mixing ratio of the air
In the literature studies on the selection of an optimal matching layer for the use of ultrasonic waves for gaseous medium measurement are dominant. However, relatively few studies on situations where flow exists inside the chamber, such as micro-leakage, have been reported. Therefore, in this study, basic research was carried out in the state where there was flow inside a chamber by generating micro-leakage. The results of the study were analyzed to assess the relationship between the temperature and the voltage inside the chamber. In addition, the ultrasonic envelope signal underwent curve fitting and was applied to micro-leakage detection for a pressure gas tank and verified experimentally.
Figure 1 shows a schematic diagram of the experimental apparatus used in
this study, and Table 1 shows the specifications of the experimental setup.
In this study, as shown in no. 14 of Fig. 1, the method of attaching an
ultrasonic sensor inside the chamber was used. In addition, as shown in no. 9, it was manufactured with a volume of 67 L, focusing on the capacity of
the currently used type-4 tank (tank made of nonmetallic liner and carbon
fiber composite material). In order to generate micro-leakage, the side of
the chamber was drilled to make a
Schematic diagram of experimental equipment.
Specifications of experimental equipment.
Ultrasonic waves experience a loss of energy due to scattering and absorption during the process of being transmitted through the medium and have characteristics of transmitting and reflecting in the boundary layer between media having different acoustic impedances. Accordingly, in the design of an ultrasonic sensor, the selection of a matching layer in consideration of acoustic impedance is an important factor.
Acoustic impedance refers to a property that hinders the propagation of
sound waves (Fatemi and Greenleaf, 1998). Acoustic impedance is calculated
by multiplying the density of a material by the speed of sound, given as
follows (Tone et al., 1985):
When ultrasonic waves radiate from the boundary layer of a medium, transmission and reflection occur, which are determined by the difference in impedance between two different media. The reflection coefficient and transmission coefficient are calculated by the following equation (Pedersen et al., 1982) by the acoustic impedance:
The structure of the ultrasonic sensor mainly consists of two matching layers to increase the frequency bandwidth. Therefore, the sound wave passing through the first matching layer in the PZT passes through the first matching layer again at the interface of the second matching layer and returns.
When the transmission coefficient transmitted from the PZT to the first
matching layer is
Specifications of matching layers.
Figure 2 schematically illustrates the structure of the ultrasonic sensor used in this study. The first matching layer and the second matching layer were attached in the direction of the radial plane of the PZT. The PZT generates vibration energy while repeating deformation and restoration when AC voltage is applied to both ends; thus, it generates ultrasonic waves in the direction opposite to the radiation surface as well as in the direction of the radiation surface. If ultrasonic waves generated in the opposite direction of the radiation surface are not absorbed, noise is increased due to vibration inside the sensor, making it difficult to obtain reliable results. Therefore, for the ultrasonic sensor fabricated in this study, a backing layer was constructed by stacking cork and rubber in a layer opposite to the radiation surface.
Ultrasonic sensor structure schematic.
In addition, in order to protect the PZT under external environmental conditions, epoxy, which has elasticity even after curing, was molded inside the housing of the sensor.
Factors that can indicate the uncertainty of the ultrasonic signal include
the temperature of the measurement medium, noise other than the measurement
frequency, and an inappropriate signal post-processing process. The
temperature of the measurement medium can affect the sound velocity of the
ultrasonic waves. The velocity of sound in air is given by the following
equation (Kim, 2019):
Looking at Eq. (5), it can be seen that the velocity of sound of ultrasonic
waves is affected by the temperature of the medium, and the speed of
velocity increases as the temperature increases. According to Eq. (5), it can be
seen that the velocity of sound decreases by about 0.18 % for every 1
Other factors that can indicate uncertainty are noise and signal post-processing, which will be discussed in detail in the sections below.
Ultrasonic waves pass through the medium from the transmitting sensor and
transmit wave energy to the receiving sensor. Unnecessary signals also occur
during this process. Therefore, in order to check the frequency of the
desired region without distortion, it is necessary to remove the frequency
region received as a noise signal. Therefore, in this paper, we used a
normal band-pass filter that blocks frequencies outside the range between
The purpose of the analysis was to consider all the characteristics of the
received ultrasonic sensor envelope signal, such as the maximum received
voltage, increasing area, decreasing area, and tail area. Therefore, in this
study, the envelope signal was analyzed using a Gaussian distribution, which
is mainly used to analyze the probability distribution of signal change, as
shown in Fig. 3, and the equation (Kim et al., 2019) given below is applied:
Algorithm of Gaussian distribution for envelope energy.
In order to conduct an experimental study on the applicability of measuring
the micro-leakage of high-pressure gas fuel using an ultrasonic sensor, an
experiment was conducted using a method of attaching an ultrasonic sensor to
the inside of the manufactured chamber. The experimental conditions are
listed in Table 3. Through a preliminary experiment, it was confirmed that
the PZT showed smooth reception sensitivity at a frequency of 84.7 kHz.
Based on this, the experiment was carried out by fixing the frequency at
84.7 kHz, and the micro-leakage locations were set at intervals of 140 mm
from the top of the chamber using holes drilled in the chamber. The amount
of micro-leakage was fixed at 3.81 L h
Experimental conditions.
Figure 4 shows the temperature variations of the gas inside the chamber due
to micro-leakage and the maximum voltage of the receiving ultrasonic sensor.
The lower the air temperature in the chamber is, the slower the sound
velocity is, and, as wave energy is transmitted through the vibrations of
the particles constituting the medium, the attenuation of the amount of
energy transmitted varies according to the density of the gas. The results
of the experiment reveal that the temperature inside the chamber decreased
from 24 to 22
Relation of received ultrasonic sensor maximum voltage with temperature in chamber.
This is due to the characteristics of ultrasonic waves transmitted through the vibrations of the particles: the more actively the particles constituting the medium move, the less loss of ultrasonic energy there is during the delivery process. When the gas medium is measured by a direct measurement method using an ultrasonic sensor, it was found that it reacts more sensitively to changes in internal temperature than to changes in the internal density of the chamber. Therefore, it is judged that precise measurement is difficult, and reliable results would be difficult to obtain when using an ultrasonic sensor to detect minute leakage of gaseous fuel from the viewpoint of the received voltage.
Figure 5a shows the results of curve fitting the envelope signal of the receiving ultrasonic sensor according to the micro-leakage at 140 mm of the micro-leakage by the Gaussian distribution at intervals of 10 min. The Gaussian envelope energy decreased nonlinearly with the occurrence of micro-leakage. Figure 5b depicts the change in the pressure inside the chamber and the change in phase according to the micro-leakage. The phase at the initial pressure of 5 bar was 0.371 ms, and it was 0.405 ms at the time of micro-leakage termination at 60 min. It can be seen that the phase shifted by 0.0057 ms on average. The phase shift according to the occurrence of micro-leakage and the pressure inside the chamber showed a similar tendency and changed linearly.
Phase shift of ultrasonic sensor envelope signal filtered through a Gaussian distribution for micro-leakage location of 140 mm.
Unlike the relationship between the received voltage and the temperature, the pressure inside the chamber and the filtered maximum envelope time show linear results. This is considered to be the result of using a Gaussian distribution that takes all the characteristics of the receiving ultrasonic sensor into account.
Phase shift of ultrasonic sensor envelope signal filtered through a Gaussian distribution for micro-leakage location of 280 mm.
Figure 6a shows the results of curve fitting the envelope signal of the received ultrasonic sensor according to the micro-leakage at 280 mm of the micro-leakage by the Gaussian distribution at intervals of 10 min. The Gaussian envelope energy decreased nonlinearly with the occurrence of micro-leakage, as in the case of the micro-leakage location of 140 mm. Figure 6b is a graph comparing the change in the pressure inside the chamber and the change in phase according to the micro-leakage. The phase at the initial pressure of 5 bar was 0.376, and it was 0.408 ms at the time of micro-leakage termination at 60 min. It can be seen that the phase shifted by 0.0053 ms on average.
Phase shift of ultrasonic sensor envelope signal filtered through a Gaussian distribution for micro-leakage location of 420 mm.
Figure 7a shows the results of curve fitting the envelope signal of the received ultrasonic sensor according to the micro-leakage at 420 mm of the micro-leakage by the Gaussian distribution at intervals of 10 min. The Gaussian envelope energy shows nonlinear results according to the micro-leakage, as in the previous results. Figure 7b is a graph comparing the change in the pressure inside the chamber and the change in phase according to the micro-leakage. The phase at the initial pressure of 5 bar was 0.357, and it was 0.388 ms at the time of micro-leakage termination at 60 min. It can be seen that the phase shifted by 0.00517 ms on average.
Phase shift of ultrasonic sensor envelope signal filtered through a Gaussian distribution for micro-leakage location of 560 mm.
Phase shift of ultrasonic sensor envelope signal filtered through a Gaussian distribution for micro-leakage location of 700 mm.
Figure 8a shows the results of curve fitting the envelope signal of the received ultrasonic sensor according to the micro-leakage at 560 mm of the micro-leakage by the Gaussian distribution at intervals of 10 min. The Gaussian envelope energy shows nonlinear results according to the micro-leakage, as in the previous results. Figure 8b is a graph comparing the change in the pressure inside the chamber and the change in phase according to the micro-leakage. The phase at the initial pressure of 5 bar was 0.368, and it was 0.395 ms at the time of micro-leakage termination at 60 min. It can be seen that the phase shifted by 0.0045 ms on average.
Figure 9a shows the results of curve fitting the envelope signal of the received ultrasonic sensor according to the micro-leakage at 700 mm of the micro-leakage by the Gaussian distribution at intervals of 10 min. The Gaussian envelope energy shows nonlinear results according to the micro-leakage, as in the previous results. Figure 9b is a graph comparing the change in the pressure inside the chamber and the change in phase according to the micro-leakage. The phase at the initial pressure of 5 bar was 0.367, and it was 0.400 ms at the time of micro-leakage termination at 60 min. It can be seen that the phase shifted by 0.0055 ms on average.
Experiments show that the thickness of the primary convergence layer of the ultrasonic sensor, as shown in Seo (2021), is similar to the frequency wavelength length (the minimum length in which a wave repeats the same shape at a given time), resulting in a relatively linear change in phase variation depending on the internal pressure of the high-pressure gas tank. This is attributed to stably transmitting signals, because the phase change is small when ultrasonic waves pass through the matching layer due to the thickness of the first matching layer of the ultrasonic sensor being the same as the frequency wavelength length. In addition, due to this effect, it is determined that the ultrasonic signal reacts sensitively according to the density change in the chamber.
Phase shift according to the pressure shift inside the chamber.
Huang et al. (2002) found that, when the temperature
inside a chamber changes from 24 to 22
Kim et al. (2019) compared the ultrasonic envelope signal
according to the mixing rate of
However, as a result of comparing the characteristics of the phase change according to the location of the micro-leakage, a certain tendency could not be confirmed. Therefore, it is difficult to detect the location of micro-leakage by a direct measurement method in which an ultrasonic sensor radiates ultrasonic waves into a gas medium.
In this paper, a study was conducted on the applicability of ultrasonic
sensors to the measurement of micro-leakage in gas fuel tanks. To proceed
with the research, a chamber was manufactured according to the capacity of a
high-pressure gas tank currently in wide use. The following conclusions were
obtained by detecting the received signal according to the location of the
micro-leakage using a direct measurement method:
The received voltage of the ultrasonic sensor showed a similar tendency to
the change in temperature. This is judged to be a result of the relationship
that the speed of sound decreases as the temperature decreases. Therefore,
it is concluded that reliable results cannot be obtained in detecting the
micro-leakage of gaseous fuel from the viewpoint of the received voltage. The phase change of the results of curve fitting the envelope signal
received by the ultrasonic sensor by a Gaussian distribution after the
occurrence of micro-leakage showed linear results according to the pressure
change inside the chamber. Because the thickness of the first matching layer
is close to the frequency wavelength length, when ultrasonic waves pass
through the matching layer, there is little phase deviation over time, and
this is attributed to stably transmitting and receiving signals. In addition, curve fitting using a Gaussian distribution has the advantage
of taking into account all the reception characteristics appearing at the
receiving ultrasonic sensor. Accordingly, it was possible to effectively
analyze multiple factors of the received ultrasonic signal according to the
pressure change in the chamber. Regarding the characteristics of the phase change according to the location
of the micro-leakage, it was difficult to detect the location of the
micro-leakage with the direct ultrasonic measurement method, because a
certain tendency could not be confirmed.
The results of this study revealed that it was difficult to use the received voltage for curve fitting the envelope signal of an ultrasonic sensor to a Gaussian distribution; the phase change shows linear results; thus, it is considered that it can be applied as a new micro-leakage measurement method for a high-pressure gas tank.
This study is focused on analyzing signal changes according to internal flow and temperature change by generating micro-leakage, unlike previous studies that experimented with no flow inside a chamber. In addition, a new application method utilizing a phase shift by post-processing the ultrasonic reception signal was proposed. By using the ultrasonic sensor, a different method was proposed for the measurement of micro-leakage of gaseous fuel that relied on the existing pressure sensor.
All the code used in this paper can be obtained upon request to the corresponding author.
All the data used in this paper can be obtained on request from the corresponding author.
WS did the experiments and wrote the article. SI provided ideas and reviewed the overall process for the experiment. GL reviewed and instructed the overall content of the thesis.
The contact author has declared that neither they nor their co-authors have any competing interests.
Publisher’s note: Copernicus Publications remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
This research was supported by the BB21plus funded by Busan Metropolitan City and Busan Institute for Talent & Lifelong Education (BIT).
This research was supported by the BB21plus funded by Busan Metropolitan City and Busan Institute for Talent & Lifelong Education (BIT).
This paper was edited by Jeong Hoon Ko and reviewed by Moo-Yeon Lee and one anonymous referee.