Electrohydrodynamic Thrusters: Silent Flight With Ionic Wind

The working principle: make ions, drift them, drag the air

An EHD thruster is a two-electrode capacitor-like structure open to the air. A very thin emitter (a wire or an array of fine wires, often 0.1–0.2 mm diameter) faces a blunt collector (a rod, tube, or airfoil leading edge) a few centimetres away, with a large DC voltage between them.

Because the wire is so thin, the electric field at its surface is enormous — field concentrates as E ∝ V / (r · ln(d/r)), so a small radius r gives a field far exceeding the bulk-gap average. When that surface field passes the breakdown threshold (in air, ≈ 3 MV/m at sea level, reached near the wire long before the gap breaks down), a corona discharge ignites. Electrons avalanche in the thin ionization sheath and are swept into the wire; the slow positive ions left behind drift across the gap toward the collector under the field.

The thrust does not come from the ions themselves — there are far too few. Each drifting ion undergoes many collisions with neutral N₂/O₂ molecules, transferring momentum on every hit. Because there are roughly a million neutrals for every ion, the momentum gets shared out into a bulk ionic wind that streams from emitter to collector. Newton's third law then pushes the whole device the other way. The thruster is essentially an electrostatic air pump that happens to have its outlet aimed backward.

The governing physics and a first-order thrust equation

To first order the thrust equals the electrostatic force on the space charge in the gap. The body force per unit volume on the ion cloud is f = ρ_c · E (charge density × field). Integrating over the gap, an elegant result falls out that is independent of geometry detail:

T = I · d / µ

where T is thrust (N), I is the corona current (A), d is the electrode gap (m), and µ is the ion mobility in air (≈ 2 × 10⁻⁴ m²/(V·s) for typical air ions). This says thrust scales with current and gap but is set by how fast ions drift for a given field. The current itself follows the space-charge-limited Mott–Gurney / Townsend law:

I ≈ C · µ · ε₀ · V (V − V₀) / d³

where V₀ is the corona onset voltage and ε₀ is vacuum permittivity. Combining them shows two hard truths. First, thrust-to-power ratio is T/P = d / (µ · V) — it falls as you raise voltage, so high voltage buys more total thrust but worse efficiency. Second, useful thrust density is small: a 4 cm gap at ≈40 kV and a few hundred µA per wire yields only a few newtons per square metre of electrode area, which is why practical thrusters use large stacked arrays of many wire–collector stages.

Real numbers: what a working ionic-wind stage delivers

• Onset voltage: corona begins around 20–30 kV for cm-scale gaps; below onset there is no current and no thrust at all.
• Operating voltage: practical designs run 20–50 kV DC. MIT's flight aircraft used ≈40 kV across ~4 cm stages.
• Current: tens to hundreds of µA per emitter wire; a full aircraft draws single-digit milliamps at tens of kilovolts.
• Thrust density: laboratory stages reach ≈3 N/m² for simple single stages, and stacked multi-stage arrays have demonstrated up to ~110 N/m² of collector area under optimized conditions.
• Thrust-to-power: ≈ 6–110 N/kW depending on voltage and speed — dramatically better than a chemical rocket at zero airspeed, but this figure collapses as the airflow speeds up.
• Total efficiency: the electromechanical efficiency of ionic wind is only ≈2–5%; most input power is lost heating the ionization region and as unrecovered ion kinetic energy.

For the 2018 MIT aircraft: 5 m wingspan, 2.45 kg, an onboard lightweight power converter stepping a battery up to ≈40 kV, sustaining flight at ≈4.8 m/s over ten 60 m flights. Thrust was on the order of a few newtons — enough only because the airframe was extremely light and low-drag.

Design trade-offs and the efficiency wall

Every EHD design sits on a set of coupled compromises:

• Gap size: a wider gap d raises thrust-to-power (which scales with d), and once you push the voltage up to sustain the current a wider gap yields more total thrust too — but that higher voltage is exactly its cost, and above a limit the corona transitions to a destructive spark. Practical gaps sit around 1–5 cm.
• Emitter sharpness: thinner wires lower onset voltage and improve ionization, but they erode and can melt; sub-0.1 mm wires are fragile and sag.
• Multi-staging: stacking many emitter–collector pairs in series (the "thruster ducts" MIT used) multiplies thrust per unit frontal area, at the cost of complex insulation and inter-stage arcing risk.
• High-voltage supply mass: the step-up converter and insulation can dominate the mass budget. The whole point of "no moving parts" is partly undone if the power electronics are heavy or fragile.

The fundamental ceiling is that thrust-to-power and jet speed are in tension. Ionic wind excels at moving a lot of air slowly at low airspeed, giving high static thrust-per-watt — but efficiency is proportional to 1/V, so extracting the high jet velocities needed for fast flight forces the voltage up and the efficiency down. This is the same physics that keeps a large slow propeller more efficient than a small fast jet, only far more severe.

Where it is actually used — and a common misconception

Despite the hype, EHD is not (yet) an aircraft engine. Its real, shipping applications exploit its silence, lack of moving parts, and ability to move air gently:

• Silent, fanless electronics cooling: ion-wind pumps and thin-film "ionic breeze" devices push air across hot components with no fan noise or vibration — of interest for laptops and sealed enclosures.
• Air ionizers / purifiers: the same corona that makes thrust also charges dust; ionic air cleaners are a mass-market EHD device (though ozone production is a real drawback).
• Boundary-layer flow control: dielectric-barrier-discharge (DBD) plasma actuators — close cousins using AC — energize the boundary layer on wings and turbine blades to delay separation.
• Research propulsion: MIT's V1/V2 aircraft and follow-on work explore near-silent, emission-free flight for small drones.

Misconception to avoid: ionic-wind thrust is not ion propulsion like a spacecraft's gridded ion engine, and it is not the discredited "antigravity" of Biefeld–Brown lifter lore. The lift of a hobbyist "lifter" is ordinary aerodynamic reaction from accelerated neutral air, not any exotic field force — and it needs atmosphere to work. In vacuum, with no neutrals to drag, an EHD thruster produces essentially nothing; it is fundamentally an air-breathing device.

A subtle engineering pitfall: measuring thrust and taming the corona

Two traps catch newcomers. First, thrust measurement is deceptively hard. A charged electrode in a lab attracts and repels nearby grounded surfaces, and the high-voltage supply's leads exert forces; naïve balance tests routinely report "thrust" that is really electrostatic attraction to the bench or a stray corona wind hitting a wall. Credible measurements isolate the device on a torsion balance, keep grounded objects far away, and verify the reaction is airflow, not field coupling.

Second, the corona is a knife-edge stability problem. Just below onset, nothing happens; just above the spark threshold, the discharge collapses into an arc that carbonizes electrodes, produces a loud crack, and can destroy the emitter and the power supply. The usable window between onset and breakdown is narrow and shifts with humidity, altitude, dust, and electrode wear. Practical thrusters run current-limited supplies and monitor for the transition. There is also an ozone and NOₓ byproduct — the same reactive chemistry that makes ion air-cleaners a health concern — which any enclosed or human-adjacent application must manage.

EHD ionic-wind thruster vs. a conventional electric ducted propeller, at small-aircraft scale

Metric	EHD ionic-wind thruster	Electric propeller/EDF
Moving parts	None (solid-state)	Rotor, bearings, gearbox
Acoustic signature	Near-silent (faint corona hiss)	70–100 dB blade/broadband noise
Thrust-to-power	≈ 6–110 N/kW (low at flight speed)	≈ 10–30 N/kW typical, higher at low speed via large disk
Drive voltage	20–50 kV DC (needs step-up + insulation)	Tens of V (LiPo direct)
Total efficiency	≈ 2–5%	60–80%
Thrust per frontal area	Low (~3–110 N/m²)	High (bounded by disk loading)