Development of Nanosecond Surface Discharge in “Actuator” Geometry
2
1312 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 36, NO. 4, AUGUST 2008 Development of Nanosecond Surface Discharge in “Actuator” Geometry Dmitry V. Roupassov and Andrei Yu. Starikovskii Abstract—Subsequent images of surface nanosecond barrier discharge development were obtained with nanosecond time res- olution. Intensified charge-coupled device camera gate was equal to τ =0.5 ns. The velocities of discharge propagation were ob- tained, whereas the discharge uniformity and filling ratio of the gap by plasma have been investigated. It was proved that two discharge flashes start during one voltage pulse on the leading and trailing pulse edges. Index Terms—Boundary layer separation control, plasma actu- ator, surface nanosecond pulsed discharge. T O DETERMINE spatial–temporal characteristics of the discharge, PicoStar HR-12 intensified charge-coupled de- vice (ICCD) camera was used. The camera was focused onto the top plane of the electrodes and dielectric layer (Fig. 1). The discharge propagation for both polarities of the high- voltage pulse was investigated (Figs. 2 and 3). The photos were obtained with a nanosecond exposure time ∆τ =0.5 ns. The repetition frequency of the incident high-voltage pulses was equal to 1 kHz; the pulse amplitude was 14 kV in the 50-Ω coaxial cable. The pulse voltage doubled on the high-voltage electrode due to the pulse reflection in the coaxial line. The half-width of the pulse was equal to 23 ns. The rise time of the pulse was equal to 8 ns, whereas the fall time was 15 ns. To initiate surface nanosecond barrier discharge, usual asym- metric plasma actuator geometry was used. These actuators are of great interest nowadays. Their ability to control the boundary layer separation makes them attractive for many flow control applications [1]. Both high-voltage electrode and low-voltage one were made of a copper foil with 0.1-mm thickness, 3-mm width, and 100-mm length. The low-voltage electrode was covered by a fluorocarbon film with a thickness of 0.4 mm. The scheme of the discharge gap is shown in Fig. 1. The ICCD camera intensifier was synchronized with the high- voltage pulse to obtain a series of subsequent photo images. A discharge develops from the edge of the high-voltage exposed electrode (upper part of the photo, Figs. 2 and 3) along the surface above the covered low-voltage electrode. The plasma layer was divided into several “streamer” channels. This is a result of the instability development of flat ionization front and of accumulated residuals (mainly charges and heat) from preceding discharge activity. A sequence of images was ob- tained using a pulse-by-pulse technique. For each high-voltage Manuscript received November 7, 2007; revised January 23, 2008. The authors are with the Physics of Nonequilibrium Systems Laboratory, Moscow Institute of Physics and Technology, 141700 Dolgoprudny, Moscow, Russia (e-mail: [email protected]). Digital Object Identifier 10.1109/TPS.2008.920294 Fig. 1. Discharge gap configuration. pulse, the intensifier delay was shifted by 0.5 ns relatively to the pulse start. Propagation of the cathode-directed discharge was subdivided into four stages. Streamers start from the up- per electrode during the first stage and run along the surface above the lower electrode (at 1st, 3rd, and 4th ns, see Fig. 2). The velocity of the discharge propagation during this stage was about 1 mm/ns. One can observe the emission from the whole streamer channel, not only from the leading ionization front. Then, the streamers’ length exceeds the length of the lower electrode, and their velocity becomes lower (0.3 mm/ns). Dur- ing this stage, only emission from the ionization front near the streamers’ heads takes place. This stage takes 5 ns (at 6th and 10th ns, see Fig. 2). After this stage, the “silent” phase of the discharge begins, when no sufficient discharge emission is observed (at 13th–20th ns, see Fig. 2). During the streamers’ propagation, the dielectric surface is charged. Thus, when the trailing edge of the high-voltage pulse reaches the electrode, the electrode potential becomes smaller than the potential of the dielectric surface in the discharge gap. This leads to the start of the second flash, which corresponds to the charge removal from the surface (at 22nd–37th ns, see Fig. 2). This flash is similar to the first one with the only difference. The discharge almost did not leave the covered electrode because there is no sufficient surface charge outside the gap (at 34th and 37th ns, see Fig. 2). Discharge propagates beyond the location of the covered elec- trode about 1 mm for the cathode-directed case and about 0.5 mm for the anode-directed case. This value corresponds to the e-field decreasing distance beyond of the edge of the covered electrode. Anode-directed discharge (Fig. 3) devel- ops almost the same way as cathode-directed one. The main difference from the cathode-directed discharge is the lower velocity of propagation (V ∼ 0.5 mm/ns). This leads to the 0093-3813/$25.00 © 2008 IEEE
Transcript of Development of Nanosecond Surface Discharge in “Actuator” Geometry
1312 IEEE TRANSACTIONS ON PLASMA SCIENCE, VOL. 36, NO. 4, AUGUST 2008
Development of Nanosecond SurfaceDischarge in “Actuator” Geometry
Dmitry V. Roupassov and Andrei Yu. Starikovskii
Abstract—Subsequent images of surface nanosecond barrierdischarge development were obtained with nanosecond time res-olution. Intensified charge-coupled device camera gate was equalto τ = 0.5 ns. The velocities of discharge propagation were ob-tained, whereas the discharge uniformity and filling ratio of thegap by plasma have been investigated. It was proved that twodischarge flashes start during one voltage pulse on the leading andtrailing pulse edges.
Index Terms—Boundary layer separation control, plasma actu-ator, surface nanosecond pulsed discharge.
TO DETERMINE spatial–temporal characteristics of thedischarge, PicoStar HR-12 intensified charge-coupled de-
vice (ICCD) camera was used. The camera was focused ontothe top plane of the electrodes and dielectric layer (Fig. 1).The discharge propagation for both polarities of the high-voltage pulse was investigated (Figs. 2 and 3). The photos wereobtained with a nanosecond exposure time ∆τ = 0.5 ns. Therepetition frequency of the incident high-voltage pulses wasequal to 1 kHz; the pulse amplitude was 14 kV in the 50-Ωcoaxial cable. The pulse voltage doubled on the high-voltageelectrode due to the pulse reflection in the coaxial line. Thehalf-width of the pulse was equal to 23 ns. The rise time ofthe pulse was equal to 8 ns, whereas the fall time was 15 ns.
To initiate surface nanosecond barrier discharge, usual asym-metric plasma actuator geometry was used. These actuatorsare of great interest nowadays. Their ability to control theboundary layer separation makes them attractive for manyflow control applications [1]. Both high-voltage electrode andlow-voltage one were made of a copper foil with 0.1-mmthickness, 3-mm width, and 100-mm length. The low-voltageelectrode was covered by a fluorocarbon film with a thickness of0.4 mm. The scheme of the discharge gap is shown in Fig. 1.The ICCD camera intensifier was synchronized with the high-voltage pulse to obtain a series of subsequent photo images. Adischarge develops from the edge of the high-voltage exposedelectrode (upper part of the photo, Figs. 2 and 3) along thesurface above the covered low-voltage electrode. The plasmalayer was divided into several “streamer” channels. This is aresult of the instability development of flat ionization frontand of accumulated residuals (mainly charges and heat) frompreceding discharge activity. A sequence of images was ob-tained using a pulse-by-pulse technique. For each high-voltage
Manuscript received November 7, 2007; revised January 23, 2008.The authors are with the Physics of Nonequilibrium Systems Laboratory,
Moscow Institute of Physics and Technology, 141700 Dolgoprudny, Moscow,Russia (e-mail: [email protected]).
Digital Object Identifier 10.1109/TPS.2008.920294
Fig. 1. Discharge gap configuration.
pulse, the intensifier delay was shifted by 0.5 ns relatively tothe pulse start. Propagation of the cathode-directed dischargewas subdivided into four stages. Streamers start from the up-per electrode during the first stage and run along the surfaceabove the lower electrode (at 1st, 3rd, and 4th ns, see Fig. 2).The velocity of the discharge propagation during this stagewas about 1 mm/ns. One can observe the emission from thewhole streamer channel, not only from the leading ionizationfront.
Then, the streamers’ length exceeds the length of the lowerelectrode, and their velocity becomes lower (0.3 mm/ns). Dur-ing this stage, only emission from the ionization front nearthe streamers’ heads takes place. This stage takes 5 ns (at 6thand 10th ns, see Fig. 2). After this stage, the “silent” phase ofthe discharge begins, when no sufficient discharge emission isobserved (at 13th–20th ns, see Fig. 2). During the streamers’propagation, the dielectric surface is charged. Thus, when thetrailing edge of the high-voltage pulse reaches the electrode,the electrode potential becomes smaller than the potential of thedielectric surface in the discharge gap. This leads to the start ofthe second flash, which corresponds to the charge removal fromthe surface (at 22nd–37th ns, see Fig. 2). This flash is similarto the first one with the only difference. The discharge almostdid not leave the covered electrode because there is no sufficientsurface charge outside the gap (at 34th and 37th ns, see Fig. 2).Discharge propagates beyond the location of the covered elec-trode about 1 mm for the cathode-directed case and about0.5 mm for the anode-directed case. This value correspondsto the e-field decreasing distance beyond of the edge of thecovered electrode. Anode-directed discharge (Fig. 3) devel-ops almost the same way as cathode-directed one. The maindifference from the cathode-directed discharge is the lowervelocity of propagation (V ∼ 0.5 mm/ns). This leads to the
0093-3813/$25.00 © 2008 IEEE
ROUPASSOV AND STARIKOVSKII: DEVELOPMENT OF NANOSECOND SURFACE DISCHARGE 1313
Fig. 2. Images of surface nanosecond barrier discharge development with nanosecond time resolution. Gate is equal to 0.5 ns. Voltage on the discharge gap isequal to 28 kV; half-width time of pulse τ1/2 = 23 ns. Cathode-directed discharge (exposed electrode is “+”).
Fig. 3. Images of surface nanosecond barrier discharge development with nanosecond time resolution. Gate is equal to 0.5 ns. Voltage on the discharge gap isequal to 28 kV; half-width time of pulse τ1/2 = 23 ns. Anode-directed discharge (exposed electrode is “−”).
reduction of the “silent” phase of the discharge (at 14th–20th ns,see Fig. 3). When the streamers’ length exceeds the length ofthe lower electrode, the discharge structure differs from thecathode-directed case (compare Fig. 2, at 6th ns and Fig. 3, at14th ns). The emission distribution is more homogeneous, andthe streamers’ heads are less pronounced. During the last stage(surface charge removal), the streamers’ structure is recovered(at 22nd–36th ns, see Fig. 3). As for the opposite polarity case,
the trailing edge of the pulse leads to the charge removal fromthe surface (at 42nd ns, see Fig. 3).
REFERENCES
[1] D. Roupassov, M. Nudnova, A. Nikipelov, and A. Starikovskii, “SlidingDBD for airflow control: Structure and dynamics,” presented at the 46thAIAA Aerospace Science Meeting and Exhibit, Reno, NV, Jan. 7–10, 2008,AIAA Paper AIAA-2008-1367.
![[XLS]Fluid Flow - Pipe sizing · Web viewOrifice discharge pressure Permanent Loss Orifice Diameter V1 Orifice Coefficient of Discharge β Orifice diameter ratio Delta P psi/100 ft](https://static.fdocument.org/doc/165x107/5ab412697f8b9ab7638b69b1/xlsfluid-flow-pipe-sizing-vieworifice-discharge-pressure-permanent-loss-orifice.jpg)
