Plasma Actuators: Controlling Airflow with Electricity

A plasma actuator is a device that pushes air using electricity alone, with no moving parts. A thin strip of electrodes on a surface, driven by a high voltage, ionizes a small layer of air. The electric field then drags the charged particles along, and they drag the neutral air with them through collisions. The result is a gentle jet of air, called an ionic wind or electric wind, flowing along the surface.

That gentle jet is the whole point. Wikipedia describes plasma actuators as devices "under development for active aerodynamic flow control." Engineers want to use them to keep airflow attached to wings at high angles, reduce drag, quiet noisy flows, cool electronics, and even steer small aircraft without flaps. Because plasma actuators have no hinges, motors or holes in the skin, and because they respond in microseconds, they are an appealing alternative to mechanical devices.

They also have real limitations. The airflow they produce is slow, the conversion of electrical power into air motion is very inefficient, and their performance can suffer in rain and humidity. This article explains how plasma actuators work, the main types, the numbers that describe their performance, a brief history, their applications and their limits, along with common misconceptions.

The physics in plain language

What "plasma" means here

A plasma is a gas in which some atoms or molecules have lost electrons, leaving a mixture of electrons, positive ions and neutral particles. The plasmas used in actuators are "non-thermal" or "cold": only a tiny fraction of the air is ionized, and while the electrons are very energetic, the bulk gas stays near room temperature. The faint violet glow seen over an operating actuator is light emitted by excited air molecules.

From charged particles to wind

The mechanism behind plasma actuators is electrohydrodynamics (EHD), the interaction between electric fields and fluids. It works in three steps:

1. Ionization. A strong electric field near the edge of an electrode accelerates free electrons, which collide with air molecules and knock off more electrons, creating ions.
2. Ion drift. The electric field pulls the ions toward the opposite electrode.
3. Momentum transfer. As the ions move, they collide constantly with neutral air molecules, which vastly outnumber them. Each collision transfers a little momentum. The neutral air is set in motion, forming a wall jet along the surface.

The force on the air is called the body force, because it acts throughout a volume of air rather than at a surface. Because the force is exerted directly by electrical means, it can be switched on, off or modulated almost instantly.

The same principle explains the old physics demonstration of the "electric wind" from a sharp point at high voltage. Wikipedia's ionic wind article notes that Francis Hauksbee made the earliest report of electric wind in 1709 and that B. Wilson demonstrated the recoil force of corona discharge in 1750. A 2018 study using schlieren imaging found that electrons play a larger role than negative ions during the negative-voltage part of the cycle, which is why some researchers prefer the term "electric wind."

Types of plasma actuator

Dielectric barrier discharge (DBD) actuators

The most widely used type for airflow control is the surface dielectric barrier discharge actuator, often just called a DBD actuator. A typical design has:

• An exposed electrode on the surface, connected to a high-voltage alternating current (AC) supply.
• A covered (encapsulated) electrode, grounded, sitting underneath a layer of insulating material and offset downstream from the exposed electrode.
• A dielectric layer between them, which prevents a direct arc between the electrodes.

When the voltage is high enough, plasma forms on the surface over the buried electrode, and the induced airflow travels from the exposed electrode toward the covered one.

The dielectric barrier is the key to stability. Charge builds up on the dielectric surface during each half-cycle, opposing the applied field and stopping each tiny discharge before it can grow into a hot arc. Because this charging must be reversed, DBD actuators need alternating or pulsed voltages; a steady DC voltage across a dielectric would simply charge the surface and stop.

Dielectric barrier discharges are not new. Wikipedia's DBD article notes that Ernst Werner von Siemens first reported them in 1857, when they were called "silent" discharges and used to make ozone. Today DBDs are used in ozone generators, plasma displays, excimer ultraviolet lamps and surface treatment of polymers and textiles. Their use for aerodynamic flow control is the newer development.

Corona discharge actuators

Corona actuators use electrodes without a dielectric barrier, often a thin wire or sharp edge, driven by DC voltage. The strong field near the sharp electrode creates a localized discharge. They can produce ionic wind efficiently but are more prone to transition into sparks or arcs.

Other designs

Wikipedia lists several more variants:

Power supplies vary. Wikipedia notes configurations using DC, AC or microwave microdischarges, and that performance is limited partly by the switching transistors (MOSFETs or IGBTs) in the high-voltage electronics.

Inside a DBD discharge: what happens each cycle

A DBD actuator does not produce a smooth, continuous plasma. Each AC cycle contains many tiny, short discharges, and the two halves of the cycle behave differently.

According to a 2014 review of surface AC DBD actuators by Nicolas Benard and Eric Moreau in Experiments in Fluids:

• During the positive-going half-cycle, the discharge is dominated by streamers, thin filaments of plasma that propagate along the surface. Current peaks reached up to 250 milliamperes in their example. One cited study measured a mean streamer duration of 30 nanoseconds, with about 35 picocoulombs of charge transferred per streamer.
• During the negative-going half-cycle, the discharge is more glow-like, with many small current pulses up to about 10 milliamperes.

The two halves do not contribute equally to the wind. The review cites results showing that the momentum transfer in the negative-going half-cycle can be at least five times larger than in the positive-going one.

Above a threshold of voltage and frequency, the discharge becomes filamentary, with bright, concentrated channels. At that point, extra electrical energy goes into heating and light rather than wind, and performance saturates.

The numbers: voltage, power, force and speed

Plasma actuators are often described qualitatively, but the published experimental literature gives a clear quantitative picture. The table below summarizes typical values from the Benard and Moreau review and Wikipedia's plasma actuator and DBD articles.

Wikipedia's plasma actuator article states the headline limit plainly: the maximum average speed induced by plasma actuators at atmospheric pressure, without a mechanical amplifier such as a chamber or cavity, is less than 10 metres per second.

Force per watt

Researchers often express performance as thrust per unit of electrical power. Results collected in the Benard and Moreau review include:

The Thomas data show a typical pattern: quadrupling the power raised thrust only about 2.5 times, so effectiveness falls as power rises. The review notes that velocity saturates at around 2 W/cm and that thicker dielectrics give higher mass flow rates.

How power scales with voltage

For thick dielectrics, the review gives the consumed power as proportional to the frequency multiplied by the square of the voltage above a threshold: P = K × f × (V − V₀)². For thin dielectrics, the exponent on the voltage term ranges from 2 to 3.5. Wikipedia's article describes the same behavior from the thrust side: thrust in still air increases as a power law of applied voltage until a threshold, beyond which the exponent drops and the actuator saturates, visible as the onset of filamentary discharge.

Worked examples

The following calculations are hypothetical, built from the typical values above, to give a feel for the scale of plasma actuator effects.

Example 1: power and thrust for a strip on a wing

Suppose a DBD actuator runs along 0.5 m of a wing's span at the typical 1 W/cm.

• Span in centimetres: 50 cm.
• Electrical power: 50 cm × 1 W/cm = 50 W.
• Using the Debien et al. figure of about 60 mN/m at 1 W/cm, thrust = 0.5 m × 60 mN/m = 30 mN.

Thirty millinewtons is the weight of about 3 grams. For comparison, Enloe and colleagues measured about 0.9 g of thrust for 55 W, about a third as much for similar power. The order of magnitude is consistent: tens of watts produce a force comparable to the weight of a few paper clips.

Example 2: how much of that power becomes air motion?

At a typical electromechanical efficiency of 0.1%, 50 W of electrical power yields about 0.05 W of mechanical power in the airflow. The remaining 99.9% ends up as heat in the plasma and dielectric, light, chemical changes (including ozone production), and electrical losses.

Example 3: actuator jet versus a real airflow

The effect of an actuator depends on how its jet compares with the surrounding flow. Dynamic pressure, q = ½ρv², measures the kinetic energy per unit volume of moving air. Using an air density of about 1.2 kg/m³ near sea level:

• Actuator jet at 7 m/s: q = 0.5 × 1.2 × 7² = 29.4 Pa.
• Airflow at 30 m/s (about 108 km/h): q = 0.5 × 1.2 × 30² = 540 Pa.
• Airflow at 70 m/s (about 252 km/h): q = 0.5 × 1.2 × 70² = 2,940 Pa.

At 30 m/s, the actuator jet carries about 5% of the free-stream dynamic pressure. At 70 m/s, about 1%. This is why plasma actuators do not try to overpower the main flow. Instead, they target sensitive points, such as the thin boundary layer near a separation point, where a small, well-timed nudge can change what the whole flow does.

Getting more from an actuator

Researchers have found several ways to improve performance without changing the basic principle.

Waveform choice

The shape of the voltage waveform matters. According to the Benard and Moreau review:

• Symmetric waveforms (sine and square) give larger mean thrust than asymmetric ones (positive and negative sawtooth).
• A sine wave is recommended for the best thrust per unit power, while a square wave gives more thrust if power consumption is less important.
• An optimized waveform reported by Kotsonis and Ghaemi raised thrust from about 40 mN/m with a sine wave to 58 mN/m, with efficiency gains of up to 30%.

Wikipedia adds that sinusoidal waveforms are often preferred because they simplify the power supply and reduce electromagnetic interference.

Burst (pulsed) operation

Rather than running continuously, actuators can be switched on and off in bursts at a lower frequency. In one set of measurements cited by Benard and Moreau, with a 1 kHz carrier at 18 kV and a 50% duty cycle, maximum induced velocity fell from 3 m/s (continuous) to 2.2 m/s (burst), regardless of whether bursts came at 20 Hz or 100 Hz. Mean velocity and consumed power both rose linearly with duty cycle over the 30% to 90% range tested. It took about ten AC cycles, roughly 10 milliseconds, for the jet to reach steady state.

Burst operation matters because many flows are most sensitive to disturbances at particular frequencies. Pulsing an actuator at the natural frequency of a flow instability can achieve more control with less average power. Wikipedia also notes that pulse-width modulation can adjust actuation strength instantaneously.

Electrode geometry

• Electrode placement. Wikipedia notes that placing the buried electrode closer to the dielectric surface produces higher induced velocities at a given voltage.
• Thin wire electrodes. Using a wire below about 25 µm in diameter as the exposed electrode produces a filament-free corona in both half-cycles; a 13 µm wire gave a 46% effectiveness gain.
• Serrated electrodes. Sawtooth-edged exposed electrodes improved thrust by 5% to 400%, with the largest gains at low voltage.

Multiple actuators in series

Placing several DBDs one after another along a surface can accelerate the air in stages, but the gains diminish. Studies collected in the review include:

Thomas and colleagues found that thrust from three DBDs grew by less than 300%, so force does not scale linearly with the number of actuators.

Surface temperature

Wikipedia reports that, at a constant peak-to-peak voltage, the maximum induced velocity depends directly on the dielectric surface temperature. Raising the surface temperature increases momentum flux while using slightly more energy, and changing the temperature can also shift the saturation point.

Anatomy of a laboratory actuator

Most published experiments use actuators that look surprisingly simple. According to the Benard and Moreau review, the electrodes are typically a few millimetres wide, with a horizontal gap between them ranging from zero to a few millimetres. Test actuators are often about 10 to 20 cm long in span. The dielectric is usually between 0.1 mm and a few millimetres thick. Dielectrics reported in the literature include PMMA (acrylic), Kapton (a polyimide film), Teflon, glass, ceramics and quartz. The buried electrode is sealed underneath so that the discharge forms only on the exposed side.

The supporting equipment is more substantial: a high-voltage amplifier or transformer, voltage and current probes to measure electrical power, a sensitive balance to measure thrust, and optical flow-measurement tools. The review draws on particle image velocimetry (PIV), which tracks tiny seed particles illuminated by a laser sheet, and laser Doppler velocimetry (LDV), which measures velocity at a point from the frequency shift of scattered laser light. Both methods avoid putting a probe into the electric field, which would disturb the discharge.

What the force really looks like

The thrust figures quoted for plasma actuators are averages over many AC cycles. Within each cycle, the force is far from steady. Time-resolved PIV measurements by Debien and colleagues, cited in the Benard and Moreau review, found a time-averaged horizontal electrohydrodynamic force of 34.2 mN/m, but the instantaneous force swung between extremes of about 345 mN/m and −293 mN/m during a cycle. In other words, the actuator alternately pushes hard forward and pulls back, and the useful wind is what remains after these large swings mostly cancel. The vertical component of the time-averaged force was tiny, about −0.37 mN/m.

There is also a difference between the force the plasma applies to the air and the thrust measured on a balance. Kriegseis and colleagues computed a total body force of about 25 mN/m but measured a thrust of only about 15 mN/m. The difference is the friction drag of the induced wall jet on the surface itself. The jet is pushed along the wall, and the wall pushes back. Anyone comparing published numbers needs to know whether a paper reports body force, measured thrust or velocity, because they are not interchangeable.

The actuator as an electrical load

From the power supply's point of view, a DBD actuator is mostly a capacitor: two electrodes separated by a dielectric. Wikipedia's DBD article notes that DBDs present a capacitive load with a power factor of only 0.1 to 0.3, and that a large amount of energy is stored in the electric field, so energy recovery is needed if the discharge is not driven continuously.

Worked example (hypothetical). If an actuator consumes 50 W of real power at a power factor of 0.2, the supply must deliver an apparent power of 50 ÷ 0.2 = 250 volt-amperes. Most of the current sloshes back and forth charging and discharging the actuator's capacitance each cycle without being consumed. The power electronics must be sized for this larger circulating current, which adds weight and cost, and is one reason supply design is such a large part of practical plasma actuator engineering.

Pressure and altitude

The behavior of any gas discharge depends on pressure. The relationship between breakdown voltage, gas pressure and electrode spacing is described by Paschen's law, which the DBD article cites when noting that recombination depends on gas pressure. Wikipedia's plasma actuator article points out that plasma is easier to generate at high altitude, where pressure is low, which is part of the interest in plasma flow control for supersonic and hypersonic vehicles. The Benard and Moreau review also mentions a test at a pressure of 5 Torr, well below one atmosphere (760 Torr). Lower pressure means fewer neutral molecules to push, however, so easier ionization does not automatically translate into a stronger flow effect.

How plasma actuators compare with other flow-control devices

Plasma actuators are one option among several for actively controlling a flow. The comparison below is qualitative and reflects the general characteristics described in the sources.

The table explains both the enthusiasm and the caution. Plasma actuators win on simplicity, speed and the absence of holes or hinges. They lose on raw authority. In practice, research tends to focus on situations where authority matters less than timing, such as triggering transition, delaying separation at low speeds, or exciting flow instabilities at their natural frequencies.

A brief history

The field grew rapidly from the late 1990s onward. Most of the work has been experimental and computational research at universities and government laboratories rather than commercial deployment.

Applications

Separation control on wings and blades

When air flows over a wing at a high angle of attack, the boundary layer can separate from the upper surface, causing a sudden loss of lift (a stall). A plasma actuator near the leading edge or near the separation point can add momentum to the slow air close to the surface, helping it stay attached. Wikipedia lists airfoil and forebody separation control and turbine blade separation control among the main applications. A 2018 study by Iranshahi and Mani examined DBD actuators as alternatives to mechanical high-lift devices such as slats and flaps.

Vortex generation

Conventional vortex generators are small fixed vanes that stir the boundary layer to delay separation. They add drag even when not needed. Plasma actuators can act as vortex generators without moving parts or holes in the surface and can be switched off when not needed. Serpentine actuators produce streamwise vortices, and Wikipedia notes recent reports of significant turbulent drag reduction in transitional flow.

Flight control

Wikipedia describes research into using plasma actuators on airfoils in place of mechanical control surfaces, which could reduce maintenance of linkages and hinges, a possible benefit for small uncrewed aircraft. A preliminary roll-control system placed DBD actuators on both sides of a NACA 0015 airfoil and used bang-bang (on/off) control, chosen because it is time-optimal and less sensitive to variations in plasma behavior. Serpentine actuators have been studied acting as slats, spoilers, flaps and ailerons, with claims that they can match a conventional aileron at low Reynolds numbers with low power consumption. The trade-off is the need for a high-voltage system that meets electromagnetic compatibility (EMC) requirements.

Noise control

Wikipedia describes two approaches: attenuating sound that passes through a plasma layer, and suppressing the flow structures that generate aerodynamic noise in the first place. Both tonal and broadband noise have been reduced in experiments, including cavity noise using streamwise and spanwise actuators.

Compressors and turbines

Plasma actuators have been studied for extending the stable operating range of axial compressors and for controlling flow over turbine blades. In film cooling, where cool air is bled through holes to form a protective layer over hot turbine blades, actuators can help keep the coolant attached to the surface, improving cooling effectiveness.

High-speed flows

At supersonic and hypersonic speeds, conventional control surfaces have limited authority, and plasma is easier to generate at high altitude where pressure is low. Wikipedia notes interest in plasma actuators for supersonic and hypersonic flow control. Localized arc filament actuators have also been used to control high-speed jets.

Cooling electronics

Electrostatic fluid accelerators create a secondary flow that thins the thermal and velocity boundary layers near hot components, improving forced-air cooling. Wikipedia mentions proposals for using them in mobile devices and notebook computers, where silent cooling with no moving parts is attractive.

Icing

Plasma actuators have been studied for icing control, as noted in 2017 research cited by Wikipedia. Because the discharge produces some heat at the surface, it may help delay or reduce ice accretion, although this is still a research topic.

The related idea: ionic wind propulsion

Plasma actuators push air along a surface to modify a flow. A related technology uses ionic wind to produce thrust directly. In November 2018, an MIT team led by associate professor Steven Barrett reported in Nature the first flight of an airplane propelled by ionic wind. According to MIT News, the aircraft had a 5-metre wingspan and weighed about 5 pounds. Batteries supplied 40,000 volts to positively charged wires strung beneath the wings. It flew 60 metres, the maximum length of the gymnasium used for testing, and the flight was repeated 10 times with similar performance.

The MIT plane used corona-discharge electrohydrodynamic thrust rather than surface DBD actuators, but it rests on the same physics: ions accelerated by an electric field transfer momentum to neutral air. It also illustrates the same trade-off: very high voltages and lightweight structures are needed to get useful forces.

Modeling plasma actuators

Predicting what a plasma actuator will do is difficult, because it involves plasma physics on nanosecond timescales and fluid dynamics on much longer ones. Wikipedia lists four modeling approaches, from most to least computationally expensive:

1. Monte Carlo with particle-in-cell: simulating individual charged particles.
2. Electrical modeling coupled with the Navier–Stokes equations: solving for the electric field and charge densities, then feeding the resulting force into a fluid simulation.
3. Lumped-element models coupled with Navier–Stokes: representing the actuator as an equivalent circuit to predict the plasma's extent and force.
4. Surrogate models: simplified force distributions fitted to experiments.

Most aerodynamic design work uses the cheaper models, representing the actuator as a body force applied in a small region of the flow. Control-oriented models have also been developed for specific cases such as cavity flow.

Limitations and open problems

Low velocity and force

The fundamental limitation is strength. Induced velocities stay below about 10 m/s for a single actuator at atmospheric pressure. That is enough to influence a laboratory flow or a low-speed drone wing but is small compared with the airflow over a full-size aircraft, where speeds are many times higher.

Low efficiency

With electromechanical efficiencies around 0.1%, plasma actuators convert very little of their electrical input into air motion. For applications where the actuator only needs to trigger a large-scale change in the flow, such as reattaching a separated boundary layer, low efficiency may be acceptable. For continuous thrust, it is a serious drawback.

Saturation

Increasing voltage beyond a threshold produces filamentary discharges and diminishing returns. The extra energy goes into heating rather than wind.

Weather sensitivity

Wikipedia notes that moisture, dew, drizzle, dust, water adhesion and icing have all been shown to affect actuators, which has discouraged practical use. Encouragingly, Lilley and colleagues reported in 2022 that when light rain was simulated by spraying droplets on an operating actuator, wet actuators quickly recovered their plasma glow and gradually regained thrust comparable to dry ones.

High-voltage engineering

Kilovolt supplies add weight, need insulation, and must meet electromagnetic compatibility requirements so they do not interfere with avionics. The power electronics, including the transistors, limit what waveforms and power levels can be delivered.

Durability and by-products

Dielectric barrier discharges are a standard way of generating ozone, so actuators produce ozone as a by-product. Long-lived installations also need dielectric materials and electrodes that stand up to continuous exposure to the discharge, which is one reason the choice of dielectric (PMMA, Kapton, Teflon, glass, ceramics or quartz) matters beyond its electrical properties.

Common misconceptions

"Plasma actuators are hot"

The plasma in a DBD actuator is non-thermal. Electrons are energetic, but the bulk gas stays near room temperature. Some surface heating does occur, and localized arc filament actuators are deliberately hot, but standard DBD actuators are not heaters.

"They push the air like a fan"

A plasma actuator does not create a pressure difference across a rotating blade. It applies a distributed body force through ion–neutral collisions in a thin layer of air near the surface. That is why its effect is concentrated within a few millimetres of the surface.

"They could replace jet engines"

The thrust from surface DBD actuators is measured in tens of millinewtons per metre of span. Even the MIT ionic wind aircraft, built specifically for propulsion, was a 5-pound demonstrator that flew 60 metres indoors. Plasma actuators are tools for controlling flows, not for propelling full-size aircraft.

"More voltage always means more effect"

Thrust follows a power law with voltage only up to a threshold. Beyond it, the discharge becomes filamentary and performance saturates. Effectiveness per watt also tends to fall as power rises.

"They are already flying on airliners"

Plasma actuators remain largely in the research and development stage. Wikipedia describes them as devices under development, and the practical obstacles of weather sensitivity, high-voltage systems and limited authority have so far kept them out of commercial service.

Frequently asked questions

How fast is the air from a plasma actuator?

A single DBD actuator at atmospheric pressure produces induced velocities of up to about 7 m/s, and multi-actuator arrays have reached about 11 m/s. Wikipedia gives the upper limit without mechanical amplification as less than 10 m/s.

How much power does one use?

Typically about 1 watt per centimetre of actuator length, with a range of 0.1 to a few W/cm. A one-metre strip at 1 W/cm uses about 100 W.

Why do they need such high voltages?

Air is a good insulator. Ionizing it requires a strong electric field, which in practical geometries means several kilovolts or more. Voltages from a few kV up to 50 kV are used, depending on the dielectric thickness.

Do they produce ozone?

Yes. Dielectric barrier discharges are a standard way to make ozone, and plasma actuators produce it as a by-product. This is relevant for enclosed applications.

Why use plasma actuators if they are so weak?

Because many flows are unstable, and small disturbances at the right place and time can produce large effects. Plasma actuators have no moving parts, are only a few hundred micrometres to a few millimetres thick, can be mounted flush on a surface, respond almost instantly and can be switched off when not needed.

What is the difference between a DBD actuator and a corona actuator?

A DBD actuator places an insulating layer between its electrodes, which limits each discharge and allows stable operation with AC voltage. A corona actuator has no barrier and typically uses a sharp electrode, such as a thin wire, with DC voltage. The Benard and Moreau review notes that very thin exposed wires, below about 25 µm, can give a corona-like, filament-free discharge even in a DBD configuration, blurring the line between the two.

Is it safe to experiment with plasma actuators?

They operate at kilovolt levels, which can deliver dangerous shocks, and they produce ozone, which irritates the lungs. Laboratory work uses insulated enclosures, interlocks and ventilation. They are not a casual home project.

What stops the velocity from going higher?

Two things. First, saturation: above a threshold, extra voltage produces filamentary discharges that heat the air rather than push it. Second, the force acts only in a thin layer near the surface, and the wall jet loses momentum to friction against that surface. Multi-actuator arrays add momentum in stages, but the gains diminish with each extra stage.

Why do researchers care about the negative half-cycle?

Measurements show the glow-like negative-going half-cycle transfers much more momentum than the streamer-dominated positive half, by a factor of at least five in one study. That finding has driven work on waveforms that lengthen or shape the most productive part of the cycle.

Are they the same as ion thrusters in spacecraft?

No. Spacecraft ion thrusters accelerate ions in a vacuum and expel them as exhaust. Plasma actuators and ionic wind devices work in air, using ions to drag neutral air molecules along.

Summary

Plasma actuators turn electricity directly into airflow by ionizing a thin layer of air and letting an electric field drag the ions, and with them the surrounding air, along a surface. The most common form, the surface dielectric barrier discharge, typically runs at several to tens of kilovolts and hundreds of hertz to tens of kilohertz, consumes about a watt per centimetre, and produces a wall jet of a few metres per second with tens of millinewtons of force per metre.

Those are small numbers, and plasma actuators convert only about a thousandth of their electrical power into air motion. Their value lies elsewhere: no moving parts, near-instant response, flush mounting, and the ability to nudge sensitive flows at exactly the right place and time. Since Roth and Sherman first applied the idea to airflow in 1998, researchers have demonstrated separation control, noise reduction, roll control, film cooling and electronics cooling. Whether plasma actuators move from the laboratory into routine engineering will depend on making them stronger, more efficient and more robust to weather.

Sources

• Plasma actuator (Wikipedia)
• Dielectric barrier discharge (Wikipedia)
• Ionic wind (Wikipedia)
• Benard, N. and Moreau, E. (2014), Electrical and mechanical characteristics of surface AC dielectric barrier discharge plasma actuators applied to airflow control, Experiments in Fluids 55:1846
• MIT engineers fly first-ever plane with no moving parts (MIT News, 21 November 2018)