On phenomena in ionized gases
Download 9.74 Mb. Pdf ko'rish
|
- Bu sahifa navigatsiya:
- References
- On the axial and radial streamer dynamics in dielectric barrier discharges
- Coarse-Grained Simulation Method for Turbulent Nonequilibrium Plasma Flows
- 2. Nonequilibrium plasma flow model
- 3. Variational Multiscale- n formulation
- Acknowledgements
- Investigation of streamer propagation and discharge development on dielectric surfaces
- Figure 1
- Ablated mass in high-voltage circuit breakers following the nature of electrode material
- Investigation of the RF power transfer efficiency of a planar ICP operated in Hydrogen
|
Aknowledgements The authors are grateful to Física de Plasmas: Diagnosis, Modelos y Aplicaciones (FQM-136) research group from the University of Córdoba and MINECO project MAT2016-79866-R for their technical and financial support. References [1] C. Yubero, M.S. Dimitrijevic, M.C. García and M.D. Calzada, Spectrochim. Acta B 62 (2007) 169.
[2] A.W. Ali and H.R. Giem, Phys. Rev. 140 (1965) 1044. [3] A. W. Ali and H. R. Griem. Phys. Rev. 144 (1966) 366. Topic number: 6 107
XXXIII. ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
On the axial and radial streamer dynamics in dielectric barrier discharges
H. Höft 1 , M. Kettlitz 1
1 P
The temporal development of the discharge channel in axial and radial direction was investigated in pulsed dielectric barrier discharges (DBDs) in a 1 mm gap at atmospheric pressure using an iCCD and a streak camera system accompanied by a fast electrical characterisation. The analysis of the two-dimensional DBD structure together with the axial and radial propagation revealed an increasing DBD emission diameter with rising axial propagation velocity (cathode-directed streamer). The radial dynamics are slower compared to the axial propagation, i.e. the radial expansion velocity (~10 4 m/s) is approx. two orders of magnitude lower than the maximal axial propagation velocity (~10 6 m/s). In addition, the streamer diameter is smaller than the channel of the transient glow-like discharge, which is formed after the streamer has crossed the gap.
Dielectric barrier discharges (DBDs) are a common tool to generate non-thermal plasmas at atmospheric pressure, which have a broad variety of applications [1]. Fundamental investigations on the DBD development, however, focus mainly on the axial dynamics, i.e. the propagation of the positive (cathode-directed) streamer and the subsequent transient glow phase. Therefore, this study is dedicated to both axial and radial breakdown dynamics. In addition, pulsed DBDs are proper test objects to study the radial dynamics of the streamer itself,
because there
are similar
underlying breakdown mechanisms, and the streamer diameter is directly connected to the electric field strength [2].
A single-filament DBD arrangement (double-sided, half-sphere Al 2 O
covered electrodes) with 1 mm gap was used [3]. The DBDs were driven by unipolar positive HV pulses with 10 _ kV amplitude and 10 _ kHz
repetition rate at fixed pulse width of 10 µs in 0.1 vol% O 2 in N
2 . Fast electrical, iCCD and streak camera measurements were performed to record the electrical characteristics as well as the spatio- temporal DBD development along and perpendicular to the discharge channel with sub-mm spatial and sub-ns temporal resolution.
In figure 1, the two-dimensional emission structure and the corresponding spatio-temporal development in axial and radial direction are shown for a DBD at the falling slope of the HV pulse. The axial DBD characteristics feature the cathode-directed streamer propagation (v max ~ 10
6 m/s) followed by the transient glow phase. The radial development is displayed at three positions in the gap; a different radial development is clearly visible, i.e. the slow expansion during the streamer propagation phase and the fast channel broadening after the streamer has crossed the gap (I,II). Directly in front of the cathode (III), no separation is visible, because the glow phase starts just when the streamer hit the cathode’s surface.
camera image, top left) and the corresponding spatio- temporal discharge development (streak camera images) along the axis and radially at positions (I) to (III) as indicated by grey bars in the iCCD shot.
[1] U. Kogelschatz, B. Eliasson, W. Egli, J. Phys. IV France Colloque C4 47-66 (1997). [2] G.V. Naidis, Phys. Rev. E 79 057401 (2009). [3] M. Kettlitz, H. Höft, T. Hoder, S. Reuter, K.- D. Weltmann, R. Brandenburg, J. Phys. D: Appl. Phys. 45 245201 (2012). Topic number 10 108
XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Coarse-Grained Simulation Method for Turbulent Nonequilibrium Plasma Flows
S.M. Modirkhazeni P 1 P , U J.P. Trelles UP 1 P
P 1 P
Nonequilibrium plasma flows in diverse applications often display instabilities and turbulence. The direct simulation of turbulent plasma flows is exceedingly expensive due to the large range of scales involved, which has motivated the development of coarse-grained simulation strategies. The Variational Multiscale-n (VMS
) method is presented as a consistent and complete approach for the coarse-grained simulation of turbulent nonequilibrium plasmas. The method builds on the VMS framework and does not rely on the use of empirical or model/problem-specific parameters. Preliminary results confirm that VMS
produces significantly more accurate results than VMS, comparable with state-of-the-art approaches for other types of flow problems.
Nonequilibrium plasmas are at the core of diverse applications, from materials processing to medicine. Thermodynamic nonequilibrium (NLTE) is generally a consequence of the interaction of plasma with processing media. This interaction also incites diverse types of instabilities and often leads to turbulence. The direct computational simulation of turbulent plasmas is often unfeasible due to the wide range of scales involved, which prompts the need for coarse-grained simulation approaches. Large Eddy Simulation (LES), the standard coarse- grained approach for turbulent incompressible flows, largely relies on assumptions not valid for plasmas.
The NLTE plasma flow model, described in [1], is constituted by a coupled set of two-temperature (heavy-species and electrons) fluid conservation and electromagnetic equations. The model is treated as a single system of transient-advective-diffusive- reactive equations, which allows its straightforward extension to accommodate other plasma models.
The Variational Multiscale-n (VMS
) method is a consistent and complete approach for the simulation of turbulent flows without the need for empirical or model/problem-specific parameters. VMS
is built on the VMS framework based on a variational decomposition of scales into large (grid-scale) and small (unresolvable) together with a residual-based approximation of the small scales [2]. A major challenge of VMS methods is the handling of the nonlinearity of the small scales, which VMS n
addresses by a fixed-point procedure [3]. The n in VMS n indicates “nonlinear” or the level of approximation used, i.e., from n = 0 for the classical VMS method, to an exact description for n = ∞. 4. Method validation and next steps Representative validation results are presented in Fig. 1 for the simulation of an incompressible turbulent free jet. The VMS n results show significantly improved accuracy than VMS and comparable to those by the dynamic Smagorinsky method, the state-of-the-art LES approach, but without
empirical or
model/problem-specific parameters. On-going efforts include the validation of the VMS
method with experimental data of the flow from a non-transferred arc plasma torch.
Fig. 1: Incompressible jet flow: instantaneous normalized velocity magnitude for different methods.
by NSF award PHY-1301935.
[1] J.P. Trelles, S.M. Modirkhazeni, Comput. Methods Appl. Mech. Engrg (2014) 282, 87-131. [2] S.M. Modirkhazeni, J.P. Trelles, Comput. Methods Appl. Mech. Engrg (2016) 306, 276-298. [3] S. ModirKhazeni, J. Trelles, 22 nd Int. Symp. Plasma Chemistry, 2015. 5 109 XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Investigation of streamer propagation and discharge development on dielectric surfaces
M. Kettlitz P 1 P , R. Klink UP 1,2 P , H. Höft 1 1
, R. Brandenburg P 1 P
P 1 P
P
P
Streamer propagation and discharge development on ceramic surfaces in nitrogen-oxygen gas mixtures at atmospheric pressure was investigated. It was possible to force the discharge to develop on the surface using pin electrodes attached directly to the dielectrics. The discharges were driven with unipolar square wave high voltages of 10 kV and 4.3 kHz. Ignition and discharge development on the surface were observed with ICCD and streak cameras. Images of single discharges showed a non-uniform and branched structure of discharge channels while accumulation over several events showed a propagation front rising from the electrode tip. The electrode polarity influenced the discharge dispersion and propagation velocity. Positive polarity of the metallic electrode (rising slope of the HV pulses) led to a cathode-directed streamer with higher propagation velocities than negative polarity (falling slope).
Surface dielectric barrier discharges (SDBDs) create transient non-thermal plasmas [1-3] and are considerably used e.g. for gas flow control or surface modification [1]. Propagation of the discharge being in contact to a dielectric surface is not fully understood yet, but is of importance for the application of SDBDs. To get insight in this mechanism, a pin-to-pin arrangement was used to investigate single localized SDBDs on ceramics in a nitrogen-oxygen gas mixture. 2. Experimental set-up The discharge arrangement consisted of a 1 mm thick alumina (AL 2 O 3 ) plate with two metal pin electrodes on each side creating a 3 mm discharge gap (figure 1). One electrode was covered with silicone, thus plasma was generated only at one side. The electrode arrangement was inserted in a gas tight plexiglass cell and a gas flow of 100 sccm of 0.1 vol% oxygen in nitrogen was flushed through the cell. A unipolar square wave high voltage pulse of 10 kV at 4.3 kHz with a pulse width of 10 µs drove the uncovered electrode. The covered one was grounded. Fast current and voltage probes monitored ignition and discharge development on the surface. The uncovered pin electrode was observed optically with ICCD and streak cameras.
During rising and falling slopes of the high voltage pulse, one discharge channel directly propagating along the gas-surface interface was formed. The discharges generated surface charges on the dielectrics, which led to a potential difference (i.e. polarity) to the pin electrode. The images of single discharges showed a non-uniform and branched structure of discharge channels while accumulation over several events showed the inception of the discharge at the electrode tip propagating off the tip. Velocity measurements with the streak camera showed the slowing down of the discharge front within 1 mm from the tip and afterwards a moderate one on the dielectrics. The electrode polarity influenced the discharge dispersion and propagation velocity. Positive polarity of the uncovered electrode (rising slope of the HV pulses) led to a cathode-directed streamer with higher propagation velocities (v max
≈ 5·10 5 m/s) than for negative polarity (falling slope). For negative polarity, the discharge was well localised at the electrode tip showing corona-like behaviour.
Figure 1: Photo of SDBDs on the ceramic surface.
[1] E. Moreau, R. Sosa, G. Artana J. Phys. D:
[2] H. Grosch, T. Hoder, K.-D. Weltmann, and R. Brandenburg, Eur. Phys. J. D 60, (2010) 547–553 [3] Y. Akishev, G. Aponin, A. Balakirev, M. Grushin, V. Karalnik, A. Petryakov, N. Trushkin, J.
Topic number 10 110 XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Ablated mass in high-voltage circuit breakers following the nature of electrode material
M. Courrege P , J-J. Gonzalez P , P. Freton PP
Laboratoire LAPLACE, Université Paul Sabatier – 31062 TOULOUSE cedex 9
In high-voltage circuit breakers (HVCB), the electrical arc created by the contacts opening, interacts with the device materials (PTFE walls and electrodes). The interaction of the arc with the electrodes leads to the presence of metallic vapours within the plasma due to ablation. These vapours greatly influence the behaviour of the arc as they modify the radiative transfer, as well as the plasma properties. We have been interested in electrode’s composition. Indeed, depending of the electrode’s material, the quantity of ablated metal differs changing the plasma behaviour. In this study two distinct electrode compositions: pure copper and copper-tungsten mixture are considered.
During contact opening in HVCB, the arc interacts with the surrounding walls and the electrodes. The interaction with the Teflon nozzles (by radiation and conduction) must be considered as it allows the flow of gas to increase the pressure in the heating volumes. The PTFE wall ablation is taking into account in many models of arc in HVCB. In our study, we use the Christen’s approach [1]. The second interaction which must be considered is related to the plasma with the electrodes. Due to its complexity and lack of experimental results this interaction is less considered in the literature. Its description requires the development of anode and the cathode models. In a thin layer near electrodes the plasma is out of equilibrium. Consequently, a hydrodynamic description is no longer suitable. To overcome the complexity of a kinetic approach, we chose to use Benilov’s ablation model [2] to take into account the thin layer and the ablation of electrodes. It allows determining the properties at the layer/plasma boundary without resolving the non-equilibrium area. A global description of the plasma with the electrodes interaction is realized using the @Fluent software. Depending on the current level vapours proportion change the plasma properties and the pressure increase in the heating volume.
Considering pure copper electrodes we have previously studied the influence of the presence of metallic vapours on temperature field, radiation, electrical conductivity, on the ablated mass of PTFE or on the pressure rise in the heating volumes. In order to get closer to real configuration, we propose simulation results obtained with
20%Cu–80%W electrodes. We can see in the figure 1 versus time the copper mass ablated assuming a pure copper electrode and a Cu-W mixture. In the presented cases the alternative current is I=25kA.
Figure 1: Copper quantity ablated from the electrode
Depending on the metallic vapours quantity the plasma properties are modified changing the HVCB behaviour. These results will be presented in a two dimensional (2D) configuration including pin and valve motions.
One transient turbulent 2D model is developed to describe the plasma behaviour in a HVCB. The interaction with the PTFE walls is considered (Conduction and Radiation by DOM and P1 models). The thin layers close to the electrodes are considered and the metallic vapour distribution calculated. Following the nature of the electrodes, the copper mass fraction field differs changing the plasma properties and the HVCB characteristics. These results will be presented and discussed.
[1] T. Christen. J. Phys. D: Apll. Phys. (2007). [2] M.S. Benilov, S. Jacobsson, A. Kaddani, S. Zahrai. J. Phys. D: Apll. Phys. (2001). 5 111
XXXIII ICPIG, July 9-14, 2017, Estoril/Lisbon, Portugal
Investigation of the RF power transfer efficiency of a planar ICP operated in Hydrogen
S. Briefi 1 , D. Rauner 1, 2 , U. Fantz 1, 2
1 AG Experimentelle Plasmaphysik, Universität Augsburg, 86135 Augsburg, Germany 2 Max-Planck-Institut für Plasmaphysik, Boltzmannstr. 2, 85748 Garching, Germany
RF coupling efficiencies were investigated for low pressure low temperature hydrogen plasmas generated in a planar ICP. The measurements revealed that the power transfer efficiency defined by the ratio of RF power delivered by the generator to the power absorbed by the plasma, exhibits a peak for varying pressure and increases with higher power. Furthermore, improves when the thickness of the dielectric window between the RF solenoid and the discharge chamber is reduced.
An important task for optimizing low pressure processing ICPs is improving the transfer of the RF power from the RF circuit to the plasma. The power transfer (or coupling) efficiency is defined as the ratio of the power delivered by the generator P delivered
to the power actually absorbed by the plasma P plasma . Two kinds of losses are imposed on P del lowering : losses due to ohmic heating in the RF network con- ductors and losses due to eddy currents that are induced e.g. in metallic parts of the vacuum system. In addition, operational parameters such as gas pressure, RF power and frequency but also the setup geometry influence the coupling efficiency [1]. Most investigations up to now focussed on rare gas plasmas but in many processing discharges, molecular gases are applied. Therefore, the investi- gations presented in this contribution were carried out in low pressure hydrogen discharges at pressures between 1 and 10 Pa.
Download 9.74 Mb. Do'stlaringiz bilan baham: 
|