Dielectric Barrier Discharge Plasma Interaction with the Airflow over an Airfoil

Pradeep Moise*, M Joy Thomas†, Joseph Mathew*, Kartik Venkatraman* * Department of Aerospace Engineering, † Department of Electrical Engineering, Indian Institute of Science, Bangalore, India 560012, † Email: jtm@ee.iisc.ernet.in

Abstract— The effect of various parameters on the velocity of the induced jet produced by a dielectric barrier discharge (DBD) plasma was studied experimentally. The glow discharge was created at atmospheric conditions by using a high voltage RF power supply. Flow visualization and photographic studies of the plasma were performed. The parametric investigation of the characteristics of the plasma show that the width of the plasma in the uniform glow discharge regime was an indication of the velocity induced under stagnant conditions. It was observed that the spanwise overlap of the two electrodes, dielectric thickness, voltage and frequency of the applied voltage are the major parameters that govern the velocity and the extent of plasma in the streamwise direction.

I. INTRODUCTION The lift, drag and stall characteristics of an aircraft’s wing

are the fundamental parameters that govern its performance [1]. They play a major role in governing the flight parameters including the take off/landing distances, maximum and sustained turn rates, flight ceiling and the climb and guide rates. Lift in an airfoil can be considered as the momentum change that occurs due to the deflection of air as it passes over the airfoil. Therefore, with an increase in the deflection of the flow, the lift increases. This is achieved by increasing the angle of attack of an airfoil as well as increasing the camber of an airfoil at a given angle of attack. But both these methods come with the severe drawback of flow separation.

Flow separation is the phenomenon in which the flow no longer follows the contour of the airfoil, leading to a loss of lift (stall). Many passive mechanical actuators like flaps, slats, slots, vortex generators, etc. have been employed in the past to prevent flow separation over a wing while increasing the lift, but all these devices have the disadvantages of having complicated designs, adding weight to the wing and acting as sources of airframe noise and vibration

Thus, a device that enhances the lift-drag aspects of an airfoil and also delays the flow separation without having the disadvantages of conventional mechanical actuators is of great practical use. In this regard, dielectric barrier discharge (DBD) devices have been shown to hold a lot of potential [2]. They have the advantages of not hindering the flow when not in use, having a quick response time and adding only negligible additional mass to the wing. Another favorable aspect of the DBD is the absence of hinges which reflect radar waves and

are detrimental to stealth of an aircraft. This study focuses on the parametric study of the plasma's characteristics in order to optimize the induced velocity of the DBD.

II. LIFT, DRAG AND FLOW SEPARATION IN AIRFOILS To achieve flight, it is imperative to overcome the

gravitational weight of the aircraft which acts vertically downwards. The wing of the aircraft is the main component which helps in generating an upward force known as the lift that counteracts the downward gravitational force. The lift is achieved by the asymmetric contour of the wing which generates a lower and higher pressure distribution on the upper (suction) and the lower (pressure) side of the airfoil respectively. The flow of air over the aircraft induces a force in the direction of flow known as the drag force. This detrimental force is overcome by the thrust produced by the engines of the aircraft. Therefore, it is a fundamental requirement to maximize the lift while reducing the corresponding drag to achieve better performances.

In an inviscid analysis of the flow over an airfoil, the pressure is initially favorable (decreases as we travel in the downstream direction) and the fluid accelerates, which can be inferred from the converging of streamlines indicating a velocity increase and consequently a pressure decrease (from Bernoulli's equation). In the next part of the flow the streamlines diverge as a result of the deceleration produced due to the increasing or adverse pressure gradient. The kinetic energy gained in the first part of the flow is used to overcome the adverse pressure gradient in the second part of the flow. In real life situations, the viscous forces that arise due to the no- slip condition on the surface of the airfoil tend to dissipate the kinetic energy gained in the form of heat. Therefore, for very strong adverse pressure gradients which develop at high angles of attack (relative angle between the flow direction and the chord of an airfoil), the kinetic energy of the fluid might not be sufficient to overcome the pressure gradient and this leads to the separation of the flow from the surface of the airfoil.

The separation leaves behind a 'dead' wake region that has almost nil velocity and a high pressure, resulting in the loss of lift, generally referred to as stall, and an increase in drag. Due to this, flow separation is a major hindrance to the performance of an airfoil. This constrains the usage of both wings as well as the other control surfaces (ailerons, rudder

978-1-4244-8286-3/10/$26.00 ©2010 IEEE 517

Figure 2. Plasma generated by the DBD

and elevator) with respect to a maximum angle of attack (known as the stall angle).

Figure 1. Model view of the DBD (Not to Scale)

III. DIELECTRIC BARRIER DISCHARGE

A. Background Though devices operating in the corona and arc discharge

regimes have long been used in the supersonic and hypersonic regime, since their effects are mainly a consequence of thermal effects on the flow, their usage has not been translated in to the low speed subsonic regime of incompressible flows.

The techniques of producing stable ‘cold’ plasmas at atmospheric conditions by placing a dielectric barrier in between two electrodes and by operating in a high frequency regime have been in use for over a decade. These non-thermal plasmas are characterized by weak ionization (degree of ionization ~10-6 to 10-4) and the electron temperature, Te being much higher than the temperature of the ions, Tion and neutral gas, Tgas (Te ~ 104K, Tion ~ Tgas ~ 300K) [3]. One of the major advantages of the DBD in industrial applications is that it operates in the uniform glow discharge regime and is a low power device (Power/feet span ~ tens of Watts). Also, studies have shown that they act by inducing a jet due to their asymmetry. Therefore, the use of DBD in the system might be viewed as a process for imparting momentum influx into the system and thus, is advantageous in the low- speed regime [4].

B. Construction Fig. 1 shows a two dimensional cross-section of the DBD. There are two electrodes arranged asymmetrically with a dielectric barrier in between. The bottom electrode is grounded and is covered completely by the dielectric. The top electrode is placed on the dielectric surface. There is an overlap of the order of millimeters between the top and the bottom electrode in the streamwise direction. The high voltage power source operating in the RF regime is connected to the upper exposed electrode.

C. Working When high frequency high voltage is supplied to the DBD

electrode system, the air between the top electrode and the projected image of the bottom electrode (virtual electrode) on the dielectric surface ionizes to form plasma (Fig. 2). The dielectric which has a higher breakdown voltage than the

voltage supplied acts as a barrier, preventing the flow of resistive currents from the top to the bottom electrode. The alternating nature of the power supplied prevents the charges from accumulating on one electrode, but allows them to oscillate for every cycle. This prevents the unwanted heating up of the electrodes which might lead to the breakdown of the dielectric barrier and creates a stable and uniform discharge.

The asymmetric configuration of the DBD induces a jet in the horizontal direction from the top to the bottom electrode, as shown in Fig. 1. Studies have shown that there is a typical acoustic sound produced during the operation of the DBD. This indicates that there is a strong plasma- neutral gas coupling of momentum which explains the production of the surface jet. When mounted on to the suction surface of the airfoil, the creation of this jet decreases the local pressure distribution, thus increasing the lift [4]. Therefore, even for angles of attack below the stall angle, the DBD is effective in increasing the lift. Also, at angles of attack above the stall angle when flow separation occurs at the leading/ trailing edge of the airfoil, the DBD when placed in the vicinity of separation produces a jet which acts as a momentum influx into the ‘dead’ wake region, imparting the required kinetic energy to allow the flow to follow the contour of the airfoil [5]. This increases the lift produced by decreasing the pressure distribution on the suction side of the airfoil and also eliminates the drag caused due to separation.

IV. EXPERIMENTAL SETUP AND PROCEDURE The effect of various parameters including voltage, supply

frequency, spanwise overlap of electrodes, width of the upper and lower electrodes, dielectric thickness, width of overlap etc., were studied by mounting the DBD in a quiescent room. Experiments included photographic study of plasma width and flow visualization using light weight Polystyrene particles placed downstream to the lower electrode (Fig. 3). Copper foils were used for the electrodes and Kapton tapes (of 0.2mm or 2 mil thickness) were used for the dielectric.

The following controlled experiments (photographic study and flow visualization) were performed-

• Spanwise overlap varying from 5cm to 20cm in steps of 2.5cm. Both the electrodes were 2.5cm wide with a 4mm thick dielectric (20x2 mil thick Kapton tapes).

518

• Upper electrode width varying from 5mm to 25mm in

steps of 5mm. The bottom electrode was 2.5cm wide with a 4mm thick dielectric in between. A spanwise overlap of 11cm was given.

• Lower electrode width varying from 5mm to 25mm in steps of 5mm. The top electrode was 2.5cm wide with a 4mm thick dielectric. A spanwise overlap of 15cm was used.

• Dielectric thickness varying from 1mm to 6mm in steps of 1mm. Both electrodes were 2.5cm wide with a spanwise overlap of 10cm.

• Streamwise overlap of the two electrodes varying from (-3 to +3) mm in steps of 1mm. Both electrodes were 2.5cm wide with a spanwise overlap of 10cm and a 4mm thick dielectric.

V. RESULTS AND DISCUSSION The flow visualization studies (Fig. 3) indicated that the

extent of the plasma in the streamwise direction (along the direction of flow) governed the magnitude of the velocity.

• The variation of the plasma width with the span of overlap of the upper and the lower electrode is shown in Fig. 4. The extent of the plasma decreases with an increase in the span of overlap for the uniform glow discharge regime, above which streamers start to develop and the velocity developed stops increasing leading to the heating up of the setup. After a particular thickness, the plasma starts to transition from a uniform glow discharge to arc discharge, limiting the induced velocity from increasing further.

• Fig. 5 shows that the extent of the plasma increases with an increase in the thickness of the dielectric which acts as the barrier. After a particular thickness, the plasma begins to undergo a transition from uniform glow discharge to arc discharge, limiting the induced velocity from increasing further.

(a) Before plasma

(b) After plasma

Figure 3. Flow visualization

• The width of the upper (exposed) electrode did not affect the plasma width, but smaller widths caused a weak discharge in the direction opposing the stream. This induced a small velocity in the other direction, which is detrimental to the objective of preventing flow separation. Insulating the upstream part of the upper electrode reduced the plasma formation in the upstream direction.

• The width of the lower (covered) electrode did not affect the plasma extent but smaller widths caused a transition to the arc regime when the width of the plasma exceeded the electrode's width, leading to the heating up of the setup.

• It was observed that a streamwise overlap of the order of a few millimeters is required to maintain a uniform glow discharge. Spacing between the exposed and the covered electrode in the streamwise direction leads to the development of a highly non-uniform plasma and heating up of the DBD setup with time.

The increase in plasma width in case of the spanwise overlap can be explained by the fact that the power is distributed along the span of the DBD. With a decrease in the span of overlap, the energy available per unit span increases. Increase in overlap is accompanied by an increase in both the visual and the acoustic intensity of the plasma. In all the cases where the transition from the glow to the arc discharge occurred, there were localized intense micro-discharges that were accompanied by a characteristic sound. Post-experimental examination of the setup indicated the heating up of the components of the DBD. This indicated that there was unwanted energy dissipation in the transition and further change of the parameter in concern is of no practical use in increasing the induced velocity. Table I. gives the variation of plasma width with respect to the governing parameters.

TABLE I. VARIATION OF PLASMA WIDTH

Variation of plasma width

Parameter Parameter value (mm) Width of plasma (mm)

50 12

100 8.3

150 6.6 Spanwise Overlap

200 3.1 1 2.8 2 4.8 3 6.8 4 8.3

Dielectric Thicknessa

5 11

519

(a) 7.5 cm (b) 10cm (c) 12.5cm

(d) 15 cm (e) 20 cm

Figure 4. Variation of plasma extent for various lengths of spanwise overlap

Figure 5. Variation of plasma extent for various dielectric thickness

REFERENCES

VI. CONCLUSIONS th[1] J. Anderson Jr., Fundamentals of Aerodynamics, 4 ed., Tata McGraw Hill, 2007, pp 19-21, 45-50, 362-374. The parameters governing the velocity of the jet induced

by a DBD device were studied. The flow visualization studies indicated that the velocity induced by the arrangement was related to the width of the plasma produced. The width of the plasma increased with the increase in the applied voltage and the applied frequency initially after which further increase in the parameter caused a gradual transition from glow to arc regime accompanied by the heating up of the system. Similar observations were made for the decrease in the spanwise overlap of the two electrodes and the increase in the dielectric thickness.

[2] T. C. Corke and M. L. Post, “Overview of plasma flow control: concepts, optimization, and application ,” AIAA Paper 2005-563, 2005.

[3] A. Dinklage, T. Klinger, G. Marx, L. Schweikhard, Plasma Physics: Confinement, Transport and Collective Effects, Springer-Verlag Berlin Hiedelberg, 2005, pp.95–105.

[4] T. C. Corke, E. Jumper, M. L. Post and T. McLaughlin, “Application of weakly-ionized plasmas as wing flow control devices”, AIAA Paper 2002-350, 2002.

[5] T. C. Corke and M. L. Post, “Separation control on High angle of attack airfoil using plasma actuators”, AIAA Paper 2003-1024, 2003.

520