Behavior of Pulsed Streamer Discharge in a Wire-Plate Electrode with Varied Gap Distances

Transcription

1 98 International Journal of Plasma Environmental Science & Technology, Vol.11, No.1, APRIL 17 Behavior of Pulsed Streamer Discharge in a Wire-Plate Electrode with Varied Gap Distances K. Nakamura 1, D. Wang, and T. Namihira 1 Graduate School of Science and Technology, Kumamoto University, Japan Institute of Pulsed Power Science, Kumamoto University, Japan Abstract Streamer discharge plasma, a type of non-thermal plasma, is a generation method of reactive radicals and ozone used for treatment of exhausted gas. Our laboratory has shown that nanosecond pulsed discharge with a pulse duration of 5 ns can generate ozone with the highest efficiency in the world. Also, as our previous research, it was shown that discharge electrodes with short gap distance succeeded to increase the concentration of ozone generation. Thus, the distance between electrodes is considered a very important parameter for applications using pulsed streamer discharge. However, how the distance between electrodes affects the pulsed discharge remains unclarified. In this research, propagation processes of pulsed streamer discharge in 4 wire-plate electrodes with different distances between electrodes were observed using an ICCD camera. The results show that, as the distance between electrodes was shortened from to, applied voltage decreased from 8 kv to 6.3 kv while discharge current increased from 149 A to 19 A; streamer head velocity didn t change from about 1.4 mm/ns; and streamer head density at propagation onset time showed no change, but conversely increasing when the streamer head reaches the plate electrode. Keywords Pulsed power, streamer discharge, streamer head, distance between electrodes I. INTRODUCTION Recently, mitigation of environmental problems using streamer discharge plasma, a type of non-thermal plasma has received much attention world-wide. Streamer discharge plasma is well known to be able to generate reactive radicals and ozone, and treat exhausted gas with high efficiency [1] [3]. The space with charged particle (ion), known as the streamer head, particularly contributes to radical production because it has the highest electric field during the discharge process [4]. Thus, many kinds of radicals are likely produced here. Recently, a nanosecond pulsed discharge consisting only of streamer discharge phase, can generate ozone with higher efficiency among other discharge methods [5] [8]. In addition, by shortening the distance between discharge electrodes, the concentration of ozone generation increased successfully [9]. Therefore, the distance between electrodes is considered a crucial parameter for applications using pulsed streamer discharge. However, various effects related to physical characteristics of the pulsed streamer discharge, including effects of the distance between electrodes, remain unclear. Further understanding of the characteristics of streamer discharge is of great importance in order to expand applications and improve treatment performance. In this research, effects of the distance between electrodes were investigated by observing the propagation process of a pulsed streamer discharge in wire-plate electrodes with varied gap distances using an ICCD camera. Corresponding author: Douyan Wang address: douyan@cs.kumamoto-u.ac.jp II. EXPERIMENTAL SETUP A. Experimental setup for this research Fig. 1 is a schematic diagram of experimental setup for this research. In this experiment, an ICCD (C7164-3, Hamamatsu Photonics KK, Japan) camera was used to observe streamer discharge in a wire-plate electrode and it was synchronized with the pulsed power generator. A wire electrode was a tungsten with a diameter of.3 mm, and a plate electrode was copper. The distance between electrodes was changeable from and. The electrode was installed in acrylic case and N /O (%) gas was fed into its acrylic case at flow rate of 1. L/min. As shown in Fig., in this experiment, to confirm streamer heads clearly, the streamer images were observed between 5 mm and 1 mm of the wire-plate electrode, though a 1 m length electrode was employed. The pulsed power generator with pulse duration of 1 ns was used to occur the discharge in the electrode. In this experiment, the exposure time of the ICCD camera was fixed at 1 ns. Applied voltage to the wire electrode was measured by a capacitive voltage divider. The current in the electrode was measured by current transformer (Current transformer, Model CT-F.5, Berogz instrumentation, USA). A digital oscilloscope (DPO754, Tektronix, USA) recorded the voltage and current signals. B. Pulsed power generator Fig. 3 shows the detail of the pulsed power generator used in this experiment. The pulsed power generator basically consisted of three Blumlein lines and a pulse transformer with a winding ratio of 1 to 3. Single Blumlein line was made by two coaxial cables (RG-31-/U, Mitsubishi Cable Industries, Japan) with characteristics impedance of 5 Ω and its total characteristic impedance was calculated at 1 Ω. In this experiment, three Blumlein lines connected in parallel at the primary side of a pulsed transformer. Thus, total characteristic

2 Nakamura et al. 99 Fig. 1. Setup schematic. Fig. 4. Waveforms without discharge under electrode. Fig.. Observation area in a wire-plate electrode. Fig. 5. Waveforms of applied voltage and discharge current. of the discharge and displacement currents. Therefore, calculating the discharge current (Idischarge ) requires subtracting the displacement current (Idisplacement ) calculated by Eq. (3). As shown in Eq. (3), Vmeasurement means the applied voltage to a wire electrode with discharge, while Celectrode means the capacitance of a wire-plate electrode obtained by Eq. (4). Fig. 4 shows waveforms of voltage v and current i shown in Eq. (4) at electrode without discharge. The capacitances respectively resulted in 1. pf 1.8 pf, 1.7 pf and 1.4 pf at,, and electrode. Finally, Fig. 5 shows waveforms of measurement current and discharge current which is subtracted displacement current at electrode. Fig. 3. Pulsed power generator with characteristic impedance of 3 Ω. impedance of the pulsed power generator calculated by Eq. (1) resulted in (3/1) (1/3) = 3 Ω. Charging voltage into the pulsed power generator was fixed at kv. Z = (N /N1 ) Z1 Idischarge = Imeasurement Idisplacement dvmeasurement Idisplacement = Celectrode dt Z 1 i dt Celectrode = v () (3) (4) (1) A. Waveform analysis III. E XPERIMENTAL RESULTS The current value, measured by a current monitor in this study, was Imeasurement as shown in Eq. (), and is the sum Fig. 6 shows waveforms of applied voltages and discharge currents under varied distances between electrodes. Table I shows a summary of peak voltage, peak current and voltage rise rate under each distance between electrodes. Voltage rise

3 1 International Journal of Plasma Environmental Science & Technology, Vol.11, No.1, APRIL Discharge Current, A Applied Voltage, kv Time, ns Time, ns Fig. 6. Waveforms of applied voltage and discharge current. TABLE I D ISCHARGE PARAMETERS U NDER E ACH D ISTANCE B ETWEEN E LECTRODES Distance between electrodes First peak voltage [kv] 6.3 1% % % 8 133% First peak current [A] 19 1% % % % Voltage rise rate [kv/ns] Fig. 7. Comparison of propagation process of streamer discharge based on distance between electrodes. time was defined as the time required to raising peak voltage from 1% to 9%. Percentages reflect ratios on the basis of electrode. An electrode with longer gap distance led an increment of peak voltage, while peak current decreased. B. Imaging of streamer discharge In general, when the pulse voltage is applied to the electrode with non-uniform electric field, orbicular streamer heads, called primary streamer, are generated in the high electric field region, corresponding to a wire electrode in this paper. Subsequently, they start propagating toward the ground electrode. After reaching the ground electrode, the propagation path of streamer heads emits lights. Finally, their path grow into plasma. This plasma is called secondary streamer [1], [11]. In this experiment, primary streamer, that is to say, streamer head was targeted for observation. Fig. 7 shows a comparison of streamer discharge phases based on the distance between electrodes, observed by an ICCD camera with exposure time fixed at 1 ns. Times of each image in Fig. 7 indicate the end of exposure time from onset of applied voltage. The degree of bright strength indicates the degree of emission intensity, with each image photographed using identical gain. The images show that, under each distance between electrodes, the wire electrode began to emit light at 15 ns, and streamer heads then started propagating from the wire electrode to the plate electrode. Finally, streamer head reached the plate electrode at 3 ns, 35 ns, 4 ns, 4.5 ns respectively. Also clear is that streamer head velocity increased near the grounded plate electrode, a phenomenon known as final jump and caused by the high electric field produced between the tip of streamer head and the plate electrode [1]. Fig. 8 shows a V -I characteristics (impedances) of an applied pulsed power into the electrodes at varied gap distances. Even when the distance between Fig. 8. V -I characteristics of applied pulsed power to electrodes with varied gap distances. electrodes was changed, impedances of all electrodes remained almost identical during streamer propagation. From Figs. 7 and 8, it can be seen that discharge current begins to flow when streamer heads start propagating, finally increasing sharply after the streamer heads reach the plate electrode. This result also shows that the reason why applied voltage decreased at shorter gap distance is considered because streamer heads bridged the electrode gap earlier and the discharge phase transferred secondary streamer earlier. C. Streamer head velocity Streamer head velocity can be calculated to require propagation distance of streamer head. The propagation distance was calculated by the analysis of luminance distribution, a simplified example of which is shown in Fig. 9. Luminance of the propagation path of the streamer head between electrodes, propagating the longest distance in each image, was integrated. This analysis thus allows acquisition of luminance distribution in specified regions. Fig. 1 shows the time history of luminance distribution analysis at each time propagation image of streamer head (P.S.) and secondary streamer (S.S.) at electrode. As can be seen, the section closer to the wire electrode ( mm) has a higher emission intensity indicating the emission of the wire electrode. Luminance

4 Nakamura et al. 11 Fig. 9. Analysis method for propagation distance of streamer head. TABLE II V ELOCITY OF S TREAMER H EAD Distance between electrodes Velocity [mm/ns] % % % % distributions after 5 ns show the emission intensity decreases once because plasma channel part connecting the steamer head with the wire electrode has no emission. The maximum value to appear next is emission intensity of streamer head. In this research, emission intensity of streamer head disappears, in other words, the tip of the streamer head was defined by propagation distance of that. It was also confirmed that the emission region of the wire electrode expanded from Fig. 7 at 3 ns and emission intensity of that increased from Fig. 1. As mentioned above, the secondary streamer (S.S.) begins to propagate after primary streamer (P.S.), that is, streamer head completely bridges the electrode gap. Even though primary streamer head was propagating, secondary streamer began to propagate. In this experiment, high voltage of 8 kv was applied to the electrode with the long gap distance of 45 mm, due to these reasons, it seems that this phenomenon was possible to observe. Fig. 11 shows the time history of propagation distance of streamer head. The x-axis of Fig. 11 indicates time, while the y-axis shows the position of the streamer head from the wire electrode. Thus, the inclinations of the lines display the propagation velocity of streamer head. Table II shows a summary of propagation velocity of streamer head when the distance between electrodes is changed. The percentage values under each parameter show the ratios of increase on the basis of electrode. This paper also evaluated streamer heads which hadn t reached the plate electrode in order to control for effects of the final jump. Table II shows that streamer head velocity during propagation remains constant, even though the distance between electrodes was changed. This is considered due to voltage rise rate, known to contribute to propagation velocity [13], being almost identical under each gap distance. D. Streamer head diameter Time history of streamer head diameter was determined from ICCD images such as Fig. 7. Streamer head diameter was calculated by the analysis of luminance distribution, a simplified example of which is shown in Fig. 1(a). Lumi- Fig. 1. Time history of luminance distributions analyzed propagation image at electrode. Fig. 11. Time history of streamer head propagation distance. nance of streamer head, propagating the longest distance in each image, was integrated. This analysis allows to obtain luminance distribution shown in Fig. 1(b). The diameter of streamer head is defined by full width at half maximum (FWHM) of the X direction and the luminance distribution is averaged over the Y direction. The same definition of the diameter was also used in [14]. Fig. 13 shows that as streamer head propagated, streamer head diameter became bigger under each distance between electrodes. It seems that sufficient pulsed voltage was able to be applied to streamer head until the end of propagation. When the gap distance was shorter, streamer head diameter at propagation ending time

5 1 International Journal of Plasma Environmental Science & Technology, Vol.11, No.1, APRIL 17 Fig. 1. Definition of streamer head diameter. Fig. 14. Definition of streamer head number. Fig. 13. Streamer head diameter based on position between electrodes. Fig. 15. Streamer head density at propagation onset time and electric field strength on the wire electrode (E ) when streamer head is generated. decreased. This is considered because streamer heads reached the plate electrode earlier, thus applying time of voltage to streamer heads was short. This result indicates the electrode with longer gap distance can generate bigger streamer heads because streamer heads are applied voltage over a longer period during propagation. In addition, streamer head diameter at, electrode increased sharply near the plate electrode. This increment was considered caused by the high electric field produced between streamer head and the plate electrode [1]. E. Streamer head density Streamer head density can be obtained by propagation images shown in Fig. 7. In Fig. 14(a), the analysis which integrates luminance in the Y -axial direction allows to obtain luminance distribution in the X-axial direction shown in Fig. 14(b). Thus, a number of streamer head was defined by a number of peak in its luminance distribution. Fig. 15 shows the relationship between streamer head density at propagation onset time ( ns in Fig. 7) and electric field strength on the wire electrode (E ) when streamer head is generated. Fig. 16 also shows time history of streamer head density. In Figs. 15 and 16, smaller plots indicate values in each image, while larger plots indicate averaged values of those. E can be calculated by Eq. (5). h R V E = h R (h R) ln h+ hr R (5) The parameters, V, h, and R, refer to applied voltage, the distance between electrodes, and the radius of a wire electrode. The time when streamer head was generated can be obtained by Fig. 11, and applied voltage at that time was about 14 kv under each gap distance due to same voltage rise rates. In case of electrodes with different gap distance used in this research, electric field enhancement factors were changed from 46.9 to Fig. 15 shows that although the distance between electrodes was changed, both streamer head density at initiation and E little changed. From this result, it is suspected that E is taken part in an increment of streamer head density at initiation. In order to increase E, it is expected from Eq. (5) using a thinner wire electrode or increasing applied voltage at initiation are required. On the other hand, Fig. 16 shows that streamer head density decreased as propagating under each gap distance. Nevertheless, when the distance between electrodes was shortened, many streamer heads could reach the plate electrode without disappearing during propagation. Due to an increment of stremaer heads which can bridge the electrode gap, it is suspected discharge current increased. Thus, the electrode with short gap distance allows many

6 Nakamura et al. 13 Fig. 16. Time history of streamer head density. streamer heads to propagate between electrodes, consequently, it is considered that ozone and various radicals are produced with high efficiency due to high-density discharge. IV. CONCLUSION In this report, effects of the distance between electrodes on pulsed streamer discharge were investigated using an ICCD camera. These results are summarized as follows. When the distance between electrodes was shortened: 1) Applied voltage decreased but discharge current increased; ) Streamer head velocity were almost constant; 3) Streamer head diameter became small. 4) Both electric field strength on the wire electrode and streamer head density at propagation onset time were unchanged; 5) Streamer head density at propagation ending time increased. Importance of the distance between electrodes on plasma processing was clarified through the observation experiment. It was shown that by shortening the gap distance, density of discharge region increases. As the result, it is possible to improve efficiency of the plasma processing using pulsed streamer discharge. [5] T. Namihira, T. Tokuichi, D. Wang, S. Katsuki, and H. Akiyama, Characterization of nano-seconds pulsed streamer discharges, in 7 16th IEEE International Pulsed Power Conference, vol. 1, June 7, pp [6] T. Matsumoto, D. Wang, T. Namihira, S. Katsuki, and H. Akiyama, Performances of nanosecond pulsed discharge, Acta Physica Polonica A, vol. 115, pp , 9. [7] H. Tamaribuchi, D. Wang, T. Namihira, S. Katsuki, and H. Akiyama, Effect of pulse width on generation of ozone by pulsed streamer discharge, in 7 16th IEEE International Pulsed Power Conference, vol. 1, June 7, pp [8] S. Masuda, M. Sato, and T. Seki, High-efficiency ozonizer using traveling wave pulse voltage, IEEE Transactions on Industry Applications, vol. IA-, pp , [9] Y. Araki, System yield on ozonizer by nano-seconds pulsed discharge (in Japanese), Master s thesis, Kumamoto University, 13. [1] G. J. J. Winands, Z. Liu, A. J. M. Pemen, E. J. M. Van Heesch, and K. Yan, Analysis of streamer properties in air as function of pulse and reactor parameters by ICCD photography, Journal of Physics D: Applied Physics, vol. 41, 8. [11] R. Ono and T. Oda, Optical diagnosis of pulsed streamer discharge under atmospheric pressure, International Journal of Plasma Environmental Science and Technology, vol. 1, pp , 7. [1] H. Pépin, D. Comtois, F. Vidal, C. Y. Chien, A. Desparois, T. W. Johnston, J. Kieffer, B. La Fontaine, F. Martin, F. A. M. Rizk, C. Potvin, P. Couture, H. P. Mercure, A. Bondiou-Clergerie, P. Lalande, and I. Gallimberti, Triggering and guiding high-voltage large-scale leader discharges with sub-joule ultrashort laser pulses, Physics of Plasmas, vol. 8, pp , 1. [13] A. Komuro, R. Ono, and T. Oda, Effects of pulse voltage rise rate on velocity, diameter and radical production of an atmospheric-pressure streamer discharge, Plasma Sources Science and Technology, vol., p. 45, 13. [14] T. M. P. Briels, J. Kos, G. J. J. Winands, E. Van Veldhuizen, and U. Ebert, Positive and negative streamers in ambient air: Measuring diameter, velocity and dissipated energy, Journal of Physics D: Applied Physics, vol. 41, p. 344, 8. REFERENCES [1] T. Namihira, S. Tsukamoto, D. Wang, S. Katsuki, R. Hackam, H. Akiyama, Y. Uchida, and M. Koike, Improvement of nox removal efficiency using short-width pulsed power, IEEE Transactions on Plasma Science, vol. 8, pp ,. [] N. Takamura, T. Matsumoto, D. Wang, T. Namihira, and H. Akiyama, Ozone generation using positive- and negative-nano-seconds pulsed discharges, in 11 IEEE Pulsed Power Conference, June 11, pp [3] A. Komuro, R. Ono, and T. Oda, Behaviour of oh radicals in an atmospheric-pressure streamer discharge studied by two-dimensional numerical simulation, Journal of Physics D: Applied Physics, vol. 46, p. 1756, 13. [4] Z. Bonaventura, A. Bourdon, S. Celestin, and V. P. Pasko, Electric field determination in streamer discharges in air at atmospheric pressure, Plasma Sources Science and Technology, vol., p. 351, 11.