On phenomena in ionized gases
Download 9.74 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- Formation of Molten Metal Jets and Droplets in the Cathode Spot of Vacuum Arc Discharge
- Study of electric field distribution in helium and hydrogen DBD at lower pressures
- 4. References
- Probing internal excitation of trapped O + ( 4 S, 2 D, 2
- Investigation of magnetic sheath effect on angle of incident ion at graphite wall
- 3. Results and Discussion
4. Conclusion The experimental results show the potential of high power regime DBD in deposition of carbon based
dust particles and their
subsequent classifications as interstellar dust analogues.
This research was financially supported by Romanian Space Agency (ROSA) under the project STAR CDI ID 486/2017-2019. The POSCCE-O 2.2.1, SMIS-CSNR 13984-901, no. 257/28.09.2010 Project, CERNESIM, is gratefully acknowledged for the infrastructure used in this study. Topic number: 14 348
XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Formation of Molten Metal Jets and Droplets in the Cathode Spot of Vacuum Arc Discharge
M. Gashkov P 1 P , N. Zubarev 12 , G.A. Mesyats 12 P , I.V. Uimanov 1 P
P
P
P
P
The dynamics of molten metal during crater formation in the cathode spot of vacuum arc discharge was theoretically investigated. At the initial stage, a liquid-metal rim is formed around the crater. This process has been numerically simulated in the framework of the two-dimensional axisymmetric heat and mass transfer problem in the approximation of viscous incompressible liquid. At a more developed stage, the motion of liquid metal loses axial symmetry, which corresponds to a tendency toward jet formation. The development of azimuthal instabilities of the rim is analysed in terms of dispersion relations for surface waves. It is shown that maximum increments correspond to instability of the Rayleigh–Plateau type. Estimations of the time of formation of liquid metal jets and their probable number are obtained.
As is known, the cathode spot is a source of liquid metal jets and droplets that play an important role in the self-sustained operation of vacuum arc discharge [1]. They are formed whereas the molten metal is extruded by the pressure of explosive plasma out from craters formed on the cathode. A jet formation mechanism based on the development of azimuthal instability of the Rayleigh–Plateau (RP) type of the boundary of liquid expelled from craters has been proposed in Ref. [2]. However, a qualitative character of models used in [2] does not exclude that the Rayleigh–Taylor (RT) instability also develops, since the characteristic times of RP and RT instabilities are comparable. The main idea of the present work is to combine numerical and analytical approaches in considering hydrodynamic processes in the cathode spot cell of vacuum arc. In the 2D axisymmetric problem formulation, we have numerically simulated the formation of a liquid metal rim around the crater. At the same time, we analytically studied linear stages of the development of 3D instabilities in the rim with allowance for a change in its geometry. As can be seen from Fig. 1, the most pronounced growth of perturbations is observed for the azimuthal harmonic with n = 11 and results from development of the RP instability. The harmonic amplitude exhibits for 25 ns an almost fivefold increase, which can provide the formation of jets simultaneously with crater formation. At the same time, the RT instability ensures most rapid growth of the harmonic with n = 5, but it’s amplitude exhibits only threefold increase. The characteristic time of development of the RP instability (i.e., the time for which the surface perturbation amplitude increases by a factor of e ≈ 2.72) amounts to 14 ns, while that for the RT instability is significantly greater and reaches 21 ns. Thus, the results of our theoretical analysis with allowance for the substantially 3D character of deformations of the liquid rim lead to the conclusion that the RP instability is responsible for the formation of liquid metal jets (see also Ref. [3]). This work was supported by the RFBR (projects Nos. 16-08-00228 and 17-08-00430) and by the Presidium of the RAS (Program No. 9).
Figure 1. The results of calculations of the growth of amplitudes of azimuthal perturbations on the surface of liquid metal rim for modes with various numbers n during the time interval from t = 10 to 35 ns (A 0 and A being the initial and final values, respectively): (1, 2) upper and lower estimates of the relative amplitude growth due to development of the RP instability; (3, 4) same for development of the RT instability. [1] G.A. Mesyats. Phys. Usp . 38 (1995) 567. [2] G.A. Mesyats, N.M. Zubarev. J. Appl. Phys. 117 (2015) 043302. [3] M.A. Gashkov, N.M. Zubarev, G.A. Mesyats, I.V. Uimanov. Tech. Phys. Lett. 42 (2016) 852.
3 349 XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Study of electric field distribution in helium and hydrogen DBD at lower pressures
S. S. Ivković 1 , B. M. Obradović 1 , N. Cvetanović 2 and M.M. Kuraica 1
P 1 P
P
P
The effect of pressure, voltage, electrode gap and surface on the electric field distribution in DBD in helium and hydrogen in the pressure range 5 – 100 mbar was investigated. It was found that the type of the discharge and its characteristic axial electric field distribution strongly depends on the distance between electrodes while influence of the pressure is mostly seen as a change of the field value. It was found that discharge in helium is a subnormal-like for 1 mm gap, and for 5 mm it is a glow-like. The discharge in hydrogen changes from a Townsend-like for 1 mm gap to a glow-like for 5 mm gap.
Atmospheric dielectric barrier discharges (DBDs) have been widely studied over the last few decades mostly because of their broad application fields [1]. Ease of their implementation was demonstrated in different processes like surface modification, deposition, activation, gas purification, decontamination. Various discharge regimes have been observed and documented using high speed imaging, temporally and spatially resolved optical emission spectroscopy and simulations. In this paper we investigate DBD in helium and hydrogen in the pressure range 5 – 100 mbar. The effect of pressure, voltage, electrode gap and surface on the electrical characteristics of DBDs is investigated. We present the different regimes observed in the DBDs operation, depending on pressure observed trough the electric field distribution in the discharge. Stark polarization spectroscopy of hydrogen Balmer alpha line [2] was used for measurement. Using time-resolved spectroscopy, evolution of the electric field distribution was studied during the discharge development.
In our experiment the discharge is formed between two parallel electrodes: one metal electrode (40 × 40 mm 2 ) is covered with alumina dielectric while the other electrode is made of steel mesh and covered with pyrex glass. The distance between the barriers is set at 1, 2 and 5 mm. The discharge chamber is firstly evacuated down to 10 -2 mbar, and then the working gas is introduced up to 80 mbar pressure. The amplitude of the sine applied voltage was 1.1 and 1.75 kV at frequency of 19.7 kHz. Voltage is measured using high-voltage probe, and current is monitored using Rogowski coil. For time- space resolved measurement of emission spectra the 1–m spectrometer with ICCD detector was used.
It was found that the type of the discharge and its characteristic axial electric field distribution strongly depends on the distance between the electrodes. Influence of the pressure is mainly observed as a change in the electric field strength, while its influence on the shape of the field distribution is minor. According to the measured electric field distributions, the discharge in hydrogen for 1 mm electrode gap is in the Townsend-like mode, for 2 mm it is in subnormal-like and for 5 mm in the glow-like mode. Figure 1 shows that discharge in helium is in subnormal-like mode for 1 mm electrode gap. While measurements for 2 mm and 5 mm gap have shown that discharge is in the glow- like mode. Fig. 1: Electric field distributions for the helium DBD at different pressures. Inter electrode gap is 1 mm.
[1] U. Kogelschatz Plasma Chem. Plasma Proc.
[2] T. Wujec, H.W. Janus and W. Jelenski, J. Phys. D: Appl. Phys. 36 (2003) 868–877.
6 -0.2 0.0
0.2 0.4
0.6 0.8
1.0 0 1 2 3 4 5 6 7 8 12mbar
20mbar 40mbar
80mbar E [kV/
cm] Distance from cathode [mm] 350
XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Probing internal excitation of trapped O + ( 4 S, 2 D, 2 P) ions by reaction with N 2
R. Plašil 1 , A. Kovalenko 1 , T. D. Tran 1 , S. Rednyk 1 , Š. Roučka 1 , P. Dohnal 1 , J. Glosík 1
1 Department of Surface and Plasma Science, Faculty of Mathematics and Physics, Charles University in Prague, Prague, Czech Republic
When the atomic oxygen cation O + takes part as a reactant or a product in experimental studies, electronically exited states may play a significant role. For investigation of ion molecular reactions, a cryo-cooled radiofrequency ion trap is used. Molecular nitrogen was chosen as a monitor gas for probing a presence of long-lived excited O + ( 2 D,
2 P) in an ensemble of ions confined in the radiofrequency trap. The fraction of electronically excited ions was determined. The rate coefficient of the reaction of O + (
S), the electronic ground state, with nitrogen molecule was measured at low temperatures. The reaction rate coefficient at 61 K was evaluated as (7.5 ± 1.5)×10 –12
cm 3 s –1 . It confirms an increase of the coefficient at low temperatures.
The atomic oxygen cation O + plays a significant role in numerous environments such as planetary ionospheres or technological plasmas. The existence of long-lived O + ( 2 D, 2 P) metastable states may be important in many situations. For states 2 D
and 2 D 5/2 the lifetimes are in the order of hours and for states 2
1/2 and
2 P 3/2 in seconds [1]. The excited states carry electronic energy of 3.3 eV and 5.0 eV, respectively. The only exoergic channel for reaction of O
+ in ground state with N 2 is formation of NO +
(1). The charge transfer (2) is endoergic. O + ( 4 S) + N
2 → NO
+ + N, ΔH = –1.13 eV (1) O +
4 S) + N
2 → N
2 + + O, ΔH = +1.96 eV (2) For the excited states O + ( 2 D, 2 P) both reactions are exoergic and for reaction (2) the rate coefficient is close to Langevin rate in the order of 10 –10
cm 3 s –1
[1, 2]. The charge transfer (2) from excited O + ion to
N 2 is dominant channel. We used these reactions as a probe for excited states of O + ions confined in the radiofrequency trap. In addition we confirmed steep increase of the rate coefficient of the reaction (2) with decreasing temperature [1, 3].
A 22-pole radiofrequency ion trap was used for this study. It was placed on a cryo-cooler in an ultra- high vacuum system. The measuring procedure was based on iterative filling of the trap with a well- defined number of primary ions O + . They react with N 2 and the contents of the trap were analysed after chosen times by means of a quadrupole mass spectrometer with micro-channel plate detector. Additional details may be found in reference [4] and references therein. The O + ions were produced in separated ion source by electron impact from N 2 O molecule to reduce the ratio of exited states in comparison with a production from O 2 .
2 + we may evaluate a fraction of exited states of O + as (5 ± 1)% for electron impact energy E e = 75 eV and (10 ± 2)% for 145 eV, see Figure 1. 0 100 200 300
400 0.02
0.1 1
e =75 eV
E e =145 eV O + N 2 + Σ [N 2 ] = 3.9×10 11 cm –3 ; [He] = 6.0×10 13 cm –3 N( t) /N 0 (t = 0)
NO +
Fig. 1: Evolution of normalized number of ions in rf trap at temperature 61 K. From the increase of NO + we
may determine the rate coefficient of the reaction (1) (7.5 ± 1.5)×10 –12 cm
3 s –1 . The number densities of the N 2
reactant and helium buffer gas are indicated in the figure. We thank the TU Chemnitz, the DFG, prof. D. Gerlich for lending us the apparatus and Czech Science Foundation grant 17-18067S. References [1] J. Glosík et al., J. Phys. B: Atom. Molec. Phys. 11 (1978) 3365. [2] B.R. Rowe et al., J. Chem. Phys. 73 (1980) 194. [3] J-L. Le Garrec et al., Chem. Phys. Lett. 372 (2003) 485. [4] D. Gerlich and G. Borodi, Faraday Discussions
1 351 XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Investigation of magnetic sheath effect on angle of incident ion at graphite wall
Nam-Kyun Kim P 1 P , Jaemin Song 1 , Younggil Jin 1 1 , and Gon-Ho Kim 1 P
P 1 P
The ion incident angle on an oblique plasma-facing surface can deviate from the magnetic field line, depending on the characteristics of the plasma sheath which is formed between the plasma and the surface. The characteristics of the sheath in an oblique magnetic field is investigated with respect to the magnetic field strength and the angle. A fluid model predicts that, in a weak magnetic field, the sheath structure consists of a presheath and a sheath. In a strong magnetic field, the sheath structure is composed of a collisional presheath, a magnetic presheath, and a sheath. The characteristics of each region and the ion dynamics inside the regions are also revealed. The observed ion incident angle, which is measured by a noble material probe, verifies the fluid model of the oblique magnetic sheath.
1. Background In the existence of a strong magnetic field (B-field), like in a tokamak, it is easy to consider that the ion incident angle on an oblique surface would be comparable to that of the B-field line. However, Ahedo [
1 ] suggested a magnetic sheath model that the electric field (E-field) inside the plasma-wall transition region (here, we call it the ‘transition region’) forces the ions to have the ExB drift motion. Thus, the ion motion becomes 3-D, deviated from the B-field line. However, no experimental observation that verifies the model has been carried out. Here, we have verified the Ahedo’s model by observing the ion incident angle on an oblique plasma-facing surface in magnetic fields. Some ambiguity of the model has also been corrected. The governing equation set of the model is, 2 2 0 e i e n n d U dx , ln . e B e eU n const k T , .
x n V const
ˆ B i i i x c i i k T dn d dU m V e e m dx dx n dx V V B x V . (1) Here, the E-field is assumed to be aligned along the surface normal (-x direction) and the magnetic field is lying on the xz-plane as B=B(cos , 0, sin ) (Fig. 1).
A noble ‘material probe’ has been developed to measure the ion incident angle at the probe surface. When immersing a negatively biased graphite probe to a hydrogen plasma, a bundle of nano-tip is formed on the surface (Fig. 1) along the ion incident direction, due to the physical/chemical etching by energetic incident ions. By analysing the inclined angle on SEM images, the ion angle can be measured. Experiments were carried out in both the weak and strong B-field conditions (an ECR source with < 1 kG and the KSTAR far-SOL with ~ 2 T). The angle between the probe (surface normal direction) and the
Figure 1. A schematic diagram of the system and the nano-scale tips on the material probe.
With the parameters of the weak-B ECR plasma, the model predicts that the transition region consists of two layers; a presheath and a sheath. In the presheath the ions start to flow along the -y direction. The y-directional velocity, V
, is given by,
2 2 2 1 sin
sin x s x c y ci Ti s V C V dU V B dx V C . (2)
Eq. (2) reveals that the ExB drift (1 st term on RHS) and the collisional property (2 nd term on RHS) both affect the ion motion. In the sheath, the ions are accelerated only in the surface normal direction. With the parameters of KSTAR far-SOL plasma, it is predicted that the transition region consists of three layers; a collisional presheath, a magnetic presheath, and a sheath. Although there is a weak E-field inside the collisional presheath, the strong B-field confines the ions thus V y =0 in the region. The ions start to move along the -y direction inside the magnetic presheath, and the velocity, 2 2
2 1 sin x s y Ti s V C dU V B dx V C ,
(3)
reveals that only the ExB drift affects the ion motion. The above effects of the sheath make the ion incident angle much narrower than the B-field angle. These prediction is greatly supported by our experiments and the details will be discussed in the conference.
[1] E. Ahedo, Phys. Plasmas 4(12), 4419 (1997). 352 |
ma'muriyatiga murojaat qiling