RESEARCH ARTICLE | JULY 20 2017

Breakdown conditions of short air-insulated gaps under

alternating non-uniform electric fields 
Yong Chen; Yuesheng Zheng; Xiren Miao

J. Appl. Phys. 122, 033304 (2017)

https://doi.org/10.1063/1.4994646


View Export
Online Citation

12 August 2024 12:31:19

 JOURNAL OF APPLIED PHYSICS 122, 033304 (2017)

Breakdown conditions of short air-insulated gaps under alternating

non-uniform electric fields
Yong Chen, Yuesheng Zheng,a) and Xiren Miao
College of Electrical Engineering and Automation, Fuzhou University, Fuzhou 350108, Fujian Province,
China
(Received 8 May 2017; accepted 5 July 2017; published online 20 July 2017)
The breakdown conditions of short air-insulated gaps under alternating non-uniform electric fields
are investigated for providing an accurate prediction of the breakdown voltages of commonly used
electrode systems. The breakdown voltages with the tip radius of the highly stressed electrode vary-
ing between 0.1 cm and 2 cm and the gap distance ranging from 1 cm to 5 cm were experimentally
investigated, and the breakdown electric field distributions in inter-electrode gaps were analyzed
by a finite element method. An abnormal phenomenon was observed that the breakdown voltage
increased in a highly non-uniform electric field with the same gap distance. For different electrode
arrangements, the breakdown conditions could not be characterized by the relationship between the
maximum breakdown electric field at the electrode tip and the field non-uniform coefficient.
Therefore, a combined breakdown criterion was proposed by introducing a geometrical parameter
of the electrode system. The breakdown conditions can be, respectively, evaluated by the streamer
inception criterion and a fitting formula when the value of the geometric parameter is on opposite
sides of a critical value. Published by AIP Publishing. [http://dx.doi.org/10.1063/1.4994646]

I. INTRODUCTION intensity is about 30 kV/cm at gap distances within

1–10 cm.11 Hepworth et al.12 built a calculation model of the
Electrode systems under non-uniform fields have been
breakdown field strength under the rod-plane and sphere-
widely employed to produce atmospheric air plasmas, which
plane gaps by considering the tip radius of the high voltage

12 August 2024 12:31:19

can be found in many emerging applications such as decom-
electrode, but this model is only limited to slow positive
position of gaseous pollutants, surface treatment, and ozone
impulses. Until now, there is no generalized law to evaluate
production, as well as in biological and medical applica-
the breakdown field strength of short air-insulated gaps under
tions.1–3 The efficiency of the plasma source can always be
improved by increasing the applied voltage, but the gap non-uniform electric fields.
breakdown should be avoided.4 In order to ensure the opera- The streamer inception criterion has been used to predict
tion safety, an accurate prediction of the breakdown voltage the breakdown voltage of air-insulated gaps by considering
is necessary. the electric field distribution in the vicinity of a high voltage
The breakdown voltages of air gaps under non-uniform electrode. It was indicated that the gap breakdown occurred
electric fields have always been experimentally investigated, when the space charge number of the electron avalanche
typically with the point-plane and sphere-plane electrodes. Les head reaches a critical value, with the electron impact and
Renadières Group5 found that when the air-insulated gap dis- attachment being considered.13 Parekh et al.14 applied this
tances were, respectively, fixed at 5 m and 10 m, the break- criterion to evaluate the breakdown voltages of positive
down voltages were higher with the larger tip radii of the high point-plane air gaps with the tip point of 0.1 cm. It was dem-
voltage electrodes. Gallet et al.6 proposed an empirical formula onstrated that this model can also be applied to electronega-
for evaluating positive impulse breakdown voltages of the rod- tive gases and gas mixtures.15–17 However, most works are
plane gaps with the gap distance from 1 m to 23 m. Then, limited to a fixed electrode system, and whether the streamer
CRIEPI7 also proposed an empirical formula for evaluating inception criterion is suitable for different electrode arrange-
impulse breakdown voltages with the gap distance in the range ments is still unclear.
of 1–25 m, which has been accepted by International In this paper, breakdown conditions of short air-insulated
Electrotechnical Commission (IEC) standard.8 Rizk9,10 pro- gaps under alternating non-uniform electric fields have been
posed a semi-empirical formula for the gap distance in the experimentally investigated. Different electrode systems were
range of 4–25 m by combining the continuous leader inception. designed for producing electric fields of different non-
However, the above results are all suitable for long air gaps, uniform degrees calculated by a finite element method. The
and the determining laws for short air gaps are still unavailable. relationship between the maximum breakdown field and the
The breakdown field intensities also have been widely field utilization factor was analyzed. The streamer inception
analyzed for evaluating the breakdown conditions under criterion was employed to analyze the breakdown conditions,
arbitrary electrode systems. Paschen’s curve is suitable for and it was not applicable when the value of the field utiliza-
uniform field gaps, which found that the breakdown field tion factor was small. Then, a combined breakdown criterion
is proposed. In addition, the experimental influencing factors
a)
Electronic mail: yueshengzheng@fzu.edu.cn are discussed.

0021-8979/2017/122(3)/033304/7/$30.00 122, 033304-1 Published by AIP Publishing.

033304-2 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

II. EXPERIMENTAL SETUP

A. Breakdown voltage measurements
The schematic of the experimental system is shown in
Fig. 1. The alternating voltage was up to 100 kVrms with a
frequency of 50 Hz. The voltage divider was used to measure
the breakdown voltages. The high voltage electrode was con-
nected to the voltage source through a protective resistance FIG. 2. Configurations of high voltage electrodes.
and the plane electrode was grounded. The electrode system
was placed on the ground of a huge high voltage testing hall. with commercial software. An example of the two-
The electrode system was cleaned with industrial alcohol dimensional axisymmetric model and the potential distribu-
10 min before experiments. The alternating voltage was tion are shown in Fig. 3. The size of the computational
ramped up at a rate of about 1 kVrms/s until there was gap domain was set as 80 cm  130 cm. The numerical results
breakdown and the transformer was shut down. The high were kept the same when the computational domain was
voltage was applied again after 1 min of resting time. Every larger than this size. The applied voltage U on the high volt-
case was tested 20 times. The value of the breakdown volt- age electrode was set as 1 kV, and the grounded plane was set
age UB was calculated by the arithmetic mean of the 20 mea- as 0 kV. These conditions were all the same for other cases.
sured values. During the experimental processes, the air The electric field distribution was calculated firstly. The maxi-
conditions were within the temperature of 301.4 6 5.1 K, the mum electric field intensity Em in the inter-electrode gap was
pressure of 1005.7 6 4.2 hPa and the relative humidity of found at the tip of the high voltage electrode surface. If the
74.1 6 9.6%, respectively. breakdown voltage was obtained, the maximum breakdown
field EBm was equal to Em  (UB/1 kV).
B. Design of electrode system The non-uniform coefficient f and the field utilization
It is well known that the electric field distribution in the factor n are always used to describe the non-uniform degree
inter-electrode gap has a close relationship with the break- of the electric field.11,19 The non-uniform coefficient is writ-
down voltage. A series of electrode systems was designed ten as,

12 August 2024 12:31:19

for producing electric fields with different non-uniform f ¼ Em =ðU=dÞ: (2)
degrees. The configurations of high voltage electrodes from
the point to the sphere are shown in Fig. 2. The tip radii r The value of f is larger than 1 under a non-uniform electric
were 0.1 cm, 0.3 cm, 0.5 cm, 1 cm and 2 cm, respectively. field. The field utilization factor n is defined as,
The grounded electrode was a plane. The materials of all
electrodes were of stainless steel. The gap distances between n ¼ 1=f : (3)
the electrode systems d were from 1 cm to 5 cm with a step
of 1 cm. Different non-uniform electric fields were produced So, the value of n is smaller than 1 under a non-uniform elec-
by the tip radius, and the gap distance varied when the high tric field.
The relationship between the non-uniform coefficient
voltage was applied.
and the electrode system is shown in Fig. 4. It has been
The electric field distributions of different electrode sys-
found that the value of f is larger due to the longer gap
tems were numerically analyzed by a finite element
method.18 Poisson’s equation is used to solve the electric
field E, which can be described as,

ru2 ¼ q=e; E ¼ ru: (1)

Here, u is the electric potential; q is the space charge den-

sity; and e is the permittivity. The calculation was carried out

FIG. 3. A two-dimensional finite element model and potential distribution

FIG. 1. Sketch of a breakdown voltage measurement system. with r ¼ 0.5 cm and d ¼ 5 cm.
033304-3 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

6:11RH 17:6ðT  273Þ

h¼  exp : (8)
0:4615T T  30
Here, RH is the relative air humidity in percent. In the same
way, the corrected maximum breakdown field under standard
reference air conditions can be written as,

EBm0 ¼ EBm =ds : (9)

III. EXPERIMENTAL RESULTS

A. Breakdown voltages
The relationship between the breakdown voltage UB0
and the electrode system is shown in Fig. 5. With the same
tip radius, the value of UB0 increased with the increase of the
gap distance. With the same gap distance, it is generally
FIG. 4. Non-uniform coefficients of different electrode systems.
believed that the value of UB0 decreased with the decrease of
the electrode curvature radius leading to the non-uniform
distribution of the electric field. When the value of the non-
distance or the smaller tip radius. The value of f is from 1.36
uniform coefficient f is smaller than 11.22 as observed in
to 2.91 when the tip radius is fixed at 2 cm, while the value
Fig. 4, this relationship is normal. However, the value of UB0
of f varied in a wide range from 6.74 to 22.65 when the tip
increased when the value of f is larger than 11.22, corre-
radius is fixed at 0.1 cm. Therefore, the variation range of the
sponding to the gap distance of 2–5 cm with the tip radius of
non-uniform coefficient with different gap distances is wider,
0.1 cm. The breakdown voltage increased in a highly non-
when the tip radius is smaller. Within the region of our
uniform electric field. With the same gap distance which is
designed electrode systems, the value of f varied from 1.36
larger than and equal to 2 cm, there is a critical radius where

12 August 2024 12:31:19

to 22.65. the value of the breakdown voltage is minimum.

C. Correction of measured data B. Maximum breakdown fields

The breakdown voltages are influenced by air condi- The maximum breakdown fields EBm0 under different
tions.11 In order to find a general law, the experimental field utilization factors n are shown in Fig. 6. When the tip
results are corrected to the values under standard reference radii are small, the values of EBm0 decreased with the increase
air conditions according to IEC standard.20 When the gap of the values of n. However, the values of EBm0 with the tip
distance is less than 0.5 m, the correction of the humidity radii of 1 cm and 2 cm are flat around 26.98 kVrms/cm and
is not required. Therefore, the corrected breakdown volt- 30.55 kVrms/cm, respectively. Different from the breakdown
age UB0 under standard reference air conditions can be voltage, the maximum breakdown fields are monotonous with
the field utilization factor. It can be found that the values of
written as,
EBm0 are much different with different tip radii when the val-
UB0 ¼ UB =ds : (4) ues of n are the same. This indicates that the breakdown con-
ditions cannot be decided only by the field utilization factor
The relative air density d and the exponent s are given by,

d ¼ ðT0 =TÞ  ðp=p0 Þ (5)

and
8
>
< 0; for g < 0:2
s ¼ g  ðg  0:1Þ=0:8; for 0:2  g < 1 (6)
>
:
1; for g  1;

where p0 is the standard reference air pressure of 1013 hPa;

T0 is the standard reference air temperature of 293 K; p and T
are the experimental air pressure and temperature, respec-
tively. The parameter g is given by,

g ¼ U=f500  d0  d½1 þ 0:012ðh=d  11Þg; (7)

where d0 is the gap distance in m; h is the absolute humidity

estimated by, FIG. 5. Breakdown voltages of different electrode systems.
033304-4 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

FIG. 6. Relationship between the maximum breakdown field and the field
utilization factor.

or the non-uniform coefficient of the electric field for an arbi-

trary electrode system.

IV. BREAKDOWN CRITERIA

A. Streamer inception criterion FIG. 7. Flow chart for calculating the maximum breakdown field.
The streamer inception criterion has been widely used to  

12 August 2024 12:31:19

predict the breakdown voltage of air-insulated gaps. A typi- Un  Q  103
DU ¼  0:015; (13)
cal expression can be written as,13,21 103
"ð #
R where n is the iterative step. The value 0.015 in Eq. (13) is
Q ¼ exp ða  gÞdl : (10) an empirical parameter for accurate and fast convergence.
0
When the maximum breakdown field is obtained, the value
The electron ionization coefficient a and the attachment of UB is equal to that of Un.
coefficient g are given by,22 The measured and calculated maximum electric fields
are compared in Fig. 8. It can be found that the calculated

3631:7 expð167:960d=EÞ for 19:0  E=d  45:6 results agree well with the experimental data when the elec-
a=d ¼ trode tip radii are 1 cm and 2 cm, with the maximum relative
7358:3 expð200:792d=EÞ for 45:6 < E=d  182:4;
error of 2.3%. When the tip radius is 0.5 cm and the value of
(11)
n is larger than 0.25 corresponding to the gap distance
2 2
g=d ¼ 9:8648  0:54ðE=dÞ þ 1:1447  10 ðE=dÞ ; (12) shorter than 2 cm, the calculated results agree well with the

where E is in kV/cm, and a and g are in cm1. Because the

experimental data have been corrected to the standard atmo-
spheric conditions, the value of d is equal to 1. The maxi-
mum electric field intensity on the high voltage electrode
surface is located at position 0. The ionization boundary R is
the position where a ¼ g. And, l is the position on the central
axis of the inter-electrode gap from the high voltage elec-
trode tip, in cm. Q is a constant. Here, Q ¼ 103 is set as the
breakdown criterion.
The computational flow for calculating the maximum
breakdown field on the high voltage electrode surface is
shown in Fig. 7. The method of calculating the initial electric
field distribution is described in Sec. II B. When the voltage
Un is applied, the electric field distribution is obtained by the
product of the initial electric field distribution and the value
of Un. If the value of Q is not converged to 103, the applied
voltage is adjusted by a step of DU. The voltage step is
defined as, FIG. 8. Comparison of measured and calculated maximum electric fields.
033304-5 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

experimental data with the maximum relative error of 4.5%. around the fitting curve within a reasonable error range.
The value of the relative error increased with the decrease of Therefore, the breakdown field formula is of great potential
the value of n. When the tip radius is 0.3 cm, the calculated to evaluate the breakdown conditions when r2/d < 0.125 cm.
results are far away from the experimental data. Therefore, Incorporating the electric field calculation into the finite ele-
the maximum breakdown field distribution can be separated ment method described in Sec. II B, the breakdown voltage
as two regions by the dotted line as shown in Fig. 8. Within can be obtained.
the lower right region, the breakdown conditions can be accu-
rately evaluated by the streamer inception criterion. However, V. DISCUSSION
this criterion is out of work within the upper left region. The value of the breakdown voltage was calculated by
the arithmetic mean value. The coefficient of variation (CV)
B. Breakdown field formula
is introduced to investigate the variation range of breakdown
In order to accurately evaluate the breakdown conditions voltages within a series of 20 breakdown events in each
within the whole region as shown in Fig. 8, the upper left case, which is defined as,
region could be considered separately. It can be found that the sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
maximum electric field is strengthened with the decrease of 1X N
ðUBm  UB Þ2
the tip radius and the increase of the gap distance. However, N i¼1
no function can be found between EBm0 and r/d with different CV ¼  100%: (15)
UB
tip radii and gap distances. So, a geometric parameter consid-
ering the structure of the electrode system is introduced, Here, UBm is the measured breakdown voltage and N is equal
which is defined as r2/d. It indicates that the maximum break- to 20. The relationship between the CV value and the elec-
down field could be a function of the surface area of the trode system is shown in Table I. It can be found that the
electrode tip and the gap distance. The relationship between maximum value of CV is only 7.4% with the tip radius of
the maximum electric field within the upper left region and 0.3 cm and the gap distance of 1 cm. Therefore, the break-
the geometric parameter is shown in Fig. 9. It can be found down voltages under short air-insulated gaps are kept stable
that all the experimental data within the upper left region under the given experimental conditions.

12 August 2024 12:31:19

are located on the left side of the point with the value of Under the electric field with a larger value of the non-
r2/d ¼ 0.125 cm, which can be defined as the critical value. uniform coefficient f, the breakdown voltage is abnormal,
On the left side of the critical point, the maximum and the value at the same gap distance is higher as shown in
breakdown field first monotonously decreased and then Fig. 5. The streamer inception criterion is unable to explain
became stable as shown in Fig. 9. The double exponential the breakdown mechanism in this range. Since the electric
function was selected to fit the experimental data by the field distribution has a close relationship with the breakdown
method of least squares. The fitting formula is, conditions, the breakdown field distributions at the gap dis-
tance of 3 cm were analyzed by the finite element method,
EBm0 ¼ 293:7 exp ð313:3r 2 =dÞ þ 48:37 exp ð2:547r2 =dÞ: and the results are shown in Fig. 10. The electric field inten-
(14) sity is the highest at the electrode tip with a radius of 0.1 cm,
which can be expected. The area of the ionization region
The fitting correlation coefficient R2 is 0.99, which indicates
where the electric field intensity is higher than about
the validity of the fitting. The fitting curve is shown in
24.2 kV/cm (a > g) is also the largest. It indicates that a large
Fig. 9. It can be found that all the experimental data are
amount of active charge is distributed around the electrode
tip, which may hinder the arc development.
All the experiments were carried out in the high voltage
laboratory of Fuzhou University (FZU). In order to verify the
experiments, a set of experiments were carried out in the high
voltage laboratory of Xiamen University of Technology
(XMUT). The same electrode systems with different tip radii
at the gap distance of 3 cm were used. The experimental
results from different laboratories are compared in Table II. It
is found that the breakdown voltages in the two laboratories

TABLE I. CV values of different electrode systems.

CV (%) r ¼ 0.1 cm 0.3 cm 0.5 cm 1 cm 2 cm

1 cm 0.0 7.4 3.3 4.3 5.7

2 cm 5.6 0.7 0.0 2.5 2.7
d ¼ 3 cm 1.9 6.1 1.7 1.9 1.4
4 cm 2.0 2.5 2.0 2.0 2.1
FIG. 9. Relationship of the maximum breakdown field and the geometric 5 cm 1.2 3.6 3.0 1.6 0.0
parameter.
033304-6 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

FIG. 10. Breakdown field distributions at a gap distance of 3 cm.

FIG. 11. Relationship between experimental data and calculated results.
are very close to the maximum relative error of 2.07%. The
value of the breakdown voltage with the tip radius of 0.1 cm electric field in kVcm2. Replacing (11) by (16), the calcu-
is also the largest at XMUT. lated results also agree well with the experimental data when
Under the same experimental conditions, the influence of the geometric parameter is higher than the critical value. The
the electrode material on the experimental results was also value of ln Q is always within the range 5–20.26 It is sug-
investigated. The original stainless steel electrodes were gested that the relations of a, g, and Q equal to 103 in this
replaced by the copper electrodes. The experimental data and paper should be considered together.22
the calculated results by the combined breakdown criterion
with different geometric parameters are shown in Fig. 11. It VI. CONCLUSION

12 August 2024 12:31:19

can be found that the distribution of experimental points is The breakdown conditions in the air-insulated gaps with
similar to the original one. The ac breakdowns always occurred the tip radius of the highly stressed electrode varying
under voltages with positive polarity. The positive streamer between 0.1 cm and 2 cm and the gap distance ranging from
development is sustained by photoionization in front of the 1 cm to 5 cm have been investigated. An abnormal phenome-
streamer head.23 So, the breakdown process has little relation- non is found that the breakdown voltages increased with the
ship with the electrode surface conditions such as the material, decrease of the tip radius with the same gap distances when
which can be deduced from Fig. 11. The calculated results fit the value of the non-uniform coefficient f is larger than
well with the experimental data, which indicates that the com- 11.22. However, the maximum breakdown fields monotoni-
bined breakdown criterion is capable of predicting the break- cally decreased with the increase of the field utilization
down voltages of short air-insulated gaps. factor. A combined breakdown criterion is proposed by
The application of the streamer inception criterion will introducing the geometric parameter r2/d. The breakdown
be affected by the relationship between a and g, as well as conditions can be evaluated by the streamer inception crite-
the Q value. The influence of using different relations is dis- rion with Q ¼ 103 when r2/d > 0.125 cm, while the condi-
cussed.24 The application of relations (11) and (12) are valid tions can be evaluated by the breakdown field formula when
only when the value of E/d is lower than 182.4 kV/cm. A r2/d < 0.125 cm.
two-exponential relation of the electron ionization coeffi-
cient is suggested for higher fields, which is written as,25 ACKNOWLEDGMENTS

a=N ¼ 396:65  1018 exp ð1603:27  1020 =ðE=NÞÞ This work was partly supported by the National Natural
Science Foundation of China under the grant 51607041 and
þ 133:37  1018 exp ð646:77  1020 =ðE=NÞÞ; the Fujian Provincial Natural Science Foundation of China
(16) under the grant 2016J01220.

where N is the gas number density in cm3; a/N is the 1

J. Winter, R. Brandenburg, and K. Weltmann, Plasma Sources Sci.
reduced ionization coefficient in cm2 and E/N is the reduced 2
Technol. 24, 064001 (2015).
K. Weltmann and T. Woedtke, Plasma Phys. Controlled Fusion 59,
014031 (2017).
TABLE II. Experimental results from different laboratories. 3
U. Kogelschatz, Plasma Chem. Plasma Process. 23, 1 (2003).
4
Y. Zheng, B. Zhang, and J. He, Phys. Plasmas 22, 023501 (2015).
UB0@FZU UB0@XMUT Relative error 5
Les Renardières Group, Electra 23, 53 (1972).
r (cm) (kVrms/cm) (kVrms/cm) (%) 6
G. Gallet, G. Leroy, R. Lacey, and I. Kromer, IEEE Trans. Power Appar.
Syst. 94, 1989 (1975).
0.1 32.78 33.36 –1.74 7
I. Kishizima, K. Matsumoto, and Y. Watanabe, IEEE Trans. Power Appar.
0.3 19.36 19.77 2.07 Syst. 103, 1211 (1984).
8
0.5 21.08 20.86 –1.05 Insulation Co-Ordination-Part 2: Application Guide, IEC 60071-2, 1996.
9
F. A. M. Rizk, IEEE Trans. Power Delivery 4, 596 (1989).
033304-7 Chen, Zheng, and Miao J. Appl. Phys. 122, 033304 (2017)

10 18
F. A. M. Rizk, IEEE Trans. Power Delivery 4, 2187 (1989). M. Larson and F. Bengzon, The Finite Element Method: Theory,
11
E. Kuffel, W. S. Zaengl, and J. Kuffel, High-Voltage Engineering: Implementation, and Applications (Springer, Heidelberg, 2013).
19
Fundamentals (Pergamon Press, Great Britain, 1984). X. Zhao, B. Li, D. Xiao, and Y. Deng, IEEE Trans. Dielectr. Electr. Insul.
12
J. K. Hepworth, R. C. Klewe, and B. A. Tozer, J. Phys. D: Appl. Phys. 5, 24, 869 (2017).
20
730 (1972). High-Voltage Test Techniques-Part 1: General Definitions and Test
13
H. Raether, Electron Avalanches and Breakdown in Gases (Butterworths, Requirements, IEC 60060-1, 2010.
21
Great Britain, 1964). Y. Zheng, B. Zhang, and J. He, Phys. Plasmas 18, 123503 (2011).
14 22
H. Parekh, M. M. A. Salama, and K. D. Srivastava, J. Appl. Phys. 49, 107 M. Sarma and W. Janischewskyj, Proc. IEE 116, 161 (1969).
23
(1978). E. Nasser and M. Heiszler, J. Appl. Phys. 45, 3396 (1974).
15 24
W. S. Zaengl, S. Yimvuthikul, and G. Friedrich, IEEE Trans. Electr. Insul. G. Georghiou, A. Papadakis, R. Morrow, and A. Metaxas, J. Phys. D:
26, 380 (1991). Appl. Phys. 38, R303 (2005).
16 25
A. Pedersen, I. W. McAllister, G. C. Crichton, and S. Vibholm, Arch. A. P. Jovanović, M. N. Stankov, V. L. Marković, and S. N. Stamenković,
Elektrotech. 67, 395 (1984). Europhys. Lett. 104, 65001 (2013).
17 26
C. Zhuang and R. Zeng, Commun. Comput. Phys. 15, 153 (2013). G. Naidis, J. Phys. D: Appl. Phys. 38, 2211 (2005).

12 August 2024 12:31:19