This work reveals essential details of plasma-surface interaction in atmospheric air that are important for a wide range of applications, beginning from airflow control and up to the high-voltage insulation. The paper discusses experimental data characterizing dynamics of development and kinetics of energy coupling in surface dielectric barrier discharge (SDBD), atmospheric air plasmas sustained over dielectric surfaces, over a wide range of time scales. The experiments have been conducted using microsecond pulse voltage waveform of single and alternating polarities. Time-resolved discharge development and mechanisms of coupling with quiescent air are analyzed using nanosecond gate camera imaging, electrical measurements, and original surface charge sensors. The results demonstrate several new, critically important processes overlooked in previous studies. Specifically, it is shown that SDBD plasmas energy release may be significantly increased by using an optimized waveform.
Beginning from early experiments [
In SDBD plasmas powered by high-voltage nanosecond pulse waveforms EHD acceleration appears to be insignificant [
In [
The experimental test cell is shown in
The discharge diagnostics include a high voltage probe (Tektronix P6015A), a current probe (Pearson model 2877), an ICCD camera (Andor iStar, minimum gate 10 ns), and surface charge sensors (CS). The surface charge sensors (CS) [
and ICCD plasma images, it may be concluded that the presence of a surface charge sensor has a negligible effect on DBD plasma parameters. The RC time constant of the sensor/voltage probe circuit is t = RC = 6 ms. The following equation was used to recalculate the surface electric potential Up from the measured value Um (R is the resistance of the voltage probe), taking into account slow charge removal from the sensor via the high- voltage probe:
In the first experimental series (Section 3.1) the power supply (TREK 20/20C plus function generator) operates in burst mode as shown in
The second experimental series (Section 3.2) was performed with a new version of the HV amplifier TREK 20/20C HS. A custom designed mixer of signals was used to make a waveform of HV power supply with alternating polarity of the pulses in the burst. The typical waveform is shown in
The surface charge dynamics were measured by the CSs. Typical data for the charge deposition is shown in
The first effect observed is that the amplitude of the surface charge is strongly influenced by a pre-history of SDBD operation. Three modes are considered for further analysis:
1) the surface is uncharged before the burst by means of application of a conducting brush;
2) the surface is charged by a previous burst of opposite polarity with no discharge contraction at negative polarity; and
3) the surface is charged by a previous burst of opposite polarity and a discharge contraction is observed at negative polarity.
The first pulse of the burst delivers the maximal electric charge compared to the next pulses except for CS3 and CS4 for the negative polarity. In the last case some charge “bleeding” to the areas far from the HV electrode, x > 10 mm, is observed. This pattern drastically changes when the contraction of the SDBD at negative polarity takes place. The most apparent difference between surface charge density dynamics in alternating polarity mode and single polarity mode discharges is in significantly higher charge transfer to the surface in the alternating polarity discharge. There is a substantial difference in transferred charge value and dynamics in regular discharge mode (streamer type discharge at positive polarity and diffusive pattern at negative polarity) compared to contraction mode at negative polarity. The contraction is typically observed at the first pulse in the burst and only if this burst is presided by positive polarity pulses.
Several studies [
An analysis of the data collected by the charge sensors was used to quantify the effect of discharge contraction and waveform polarity on charge transfer and energy coupled. The result of these measurements is summarized in
The data processing indicates that the discharge contraction significantly increases the energy stored on the dielectric surface, which in this case may exceed the energy dissipated as Joule heat. The stored energy is dissipated if the discharge pulse is followed by an opposite polarity pulse. In a single polarity discharge, on the other hand, surface charge accumulation limits energy coupled to the plasma by subsequent pulses.
For the second experimental series with the alternating polarity of pulses the analysis of the charge deposition dynamics and distribution over the dielectric surface is fulfilled based on the data of the charge sensors.
using Equation (1). Three datasets are shown for comparison: alternating polarity (two charts, for sensors S1-S4 and for S1-S3, S5), positive polarity, and negative polarity pulse sequences. The amplitude of the supplied voltage (HV amplifier output without the test cell) is similar for all three cases, Umax = 15 kV. Note, some residual surface charge may remain on the dielectric surface originating from a previous burst regardless of surface neutralization procedure.
An essential dissimilarity between the bipolar pulse and the single polarity pulse trains is apparent. Looking at CS2-CS4 a change of the surface potential during the single polarity pulses is negligibly small compared to alternating polarity pulses, except for the first pulse. The stepwise modulation of the surface potential for a bipolar pulse train means a significant charge transfer (read: electrical current) up to Q > 1 µC with each pulse. Second, a surface area relevant to the charge deposition is bigger for the bipolar pulses than for the single polarity, especially the negative polarity pulses. The length of the charge deposition area is x < 12 mm in the case of single polarity while for the bipolar pulsing it is x > 21 mm. The third important feature of the discharge operation is a “swing” effect of the surface potential. This phenomenon appears as a result of an increase of the amplitude of charge deposition during a few first pulses, see
Detailed data on the surface charge dynamics is presented in
As it was shown in previous work [
the region close to the high-voltage electrode, and the asymptotic surface charge density is typically lower compared to regions further away from the electrode. The charge deposited on the dielectric surface further away from the electrode stays there for a longer time, and a significant fraction of it, up to s » 0.2 mC/cm2, remains there until arrival of an opposite polarity pulse. The last time scale is approximate and strongly depends on the ambient conditions, such that the residual charge can be detected on the surface after several hours.
The distribution of the electric potential over the dielectric surface is shown in
The discharge contraction, which is observed to begin on the 3rd pulse at the latest, significantly increases the surface area charged by the plasma. No less important to the energy balance is that those portions of the electric charge (deposited far from the HV electrode) are not being removed from the surface after the pulse but remain there before the pulse of opposite polarity comes.
The analysis performed above allows us to quantify the energy balance of the discharge in the considered configuration. The total energy coupled at the regular pulse (after, at least, 5 pairs of pulses) is calculated by means of a typical method [
An observation of
Parameter | Alternating polarity pulses | Single polarity pulses | ||
---|---|---|---|---|
Pulse polarity | + | − | + | − |
Total energy coupled per regular pulse, mJ | 6 ± 2 | 6 ± 2 | 1.5 ± 0.5 | 2 ± 0.5 |
Electrostatic energy at Umax, 1st pulse, mJ | 4.4 | 3.6 | 2.8 | 2.2 |
Electrostatic energy at Umax, regular pulse, mJ | 7.4 | 6.8 | 3.5 | 2.8 |
Electrostatic energy 100 µs after pulse, mJ | 3.1 | 2.7 | 0.6 | 0.2 |
the surface between the pulses but most of it dissipates with a time scale of t » 30 µs. Contrary to that, in the case of alternating polarity a significant fraction of the energy is conserved on the surface in the electrostatic form. It partially releases at an opposite polarity pulse.
The dynamics of the surface potential distribution during an individual pulse and after it is presented in
This experimental study of atmospheric SDBD examines different supplied voltage waveforms in terms of surface charge dynamics and energy release in the near-surface gas. It also extends the experimental findings of previous work [
The time-dependent distribution of the surface electric potential is measured by means of original charge sensors. It is shown that the alternating polarity of the supplied voltage gives a significant benefit in the
discharge area and the power deposition, increasing it by a factor of 2 - 4. The key factor of the discharge dynamics is the development of ionization instability that appears in the contraction of the discharge current and formation of the filamentary structure of the plasma for both positive and negative polarities. The contraction significantly increases the effective area of the electric charge deposition. A main criterion of the discharge contraction is the generation of a zone with a high level of longitudinal electric field, not less than 15 kV/cm, realized at a change of the sign of the surface charge.
The results demonstrate that surface plasma actuator control authority may be remarkably increased by using an alternating polarity pulse waveform, which is more effective than the removal of surface charge between the pulses.
This work was supported by the FlowPAC Institute, Department of Aerospace and Mechanical Engineering, University of Notre Dame. Authors express their gratitude to Professors Igor Adamovich and Victor Soloviev for multiple discussions.
Alec Houpt,Sergey B. Leonov, (2015) Dynamics of Charge Transfer by Surface Electric Discharges in Atmospheric Air. Journal of Applied Mathematics and Physics,03,1062-1071. doi: 10.4236/jamp.2015.38132