Pulse Detonation Engines: Thrust From Controlled Explosions

Deflagration vs. detonation: why the burn mode matters

Ordinary flames — in a candle, a jet combustor, or a gasoline cylinder — are deflagrations: the reaction front is a subsonic wave that spreads by conducting heat into the unburned mixture, moving at perhaps 0.4 m/s (a laminar methane flame) to a few tens of m/s. Pressure is essentially constant across a deflagration.

A detonation is a fundamentally different beast. It is a shock wave with combustion locked to its back face: the shock compresses and heats the mixture past its autoignition point in nanoseconds, the heat release feeds energy back into the shock, and the whole coupled structure propagates supersonically — typically 1,800–2,400 m/s in fuel-air mixtures. Because the shock does the compressing, a detonation raises pressure by roughly 15–20× as it passes. That self-compression is the entire point: a PDE gets a pressure rise for free that a turbojet has to pay for with a heavy multi-stage compressor.

The governing physics: Chapman-Jouguet and the Humphrey cycle

A self-sustaining detonation settles onto one unique speed — the Chapman-Jouguet (CJ) velocity — the condition where the burned gas leaves the wave at exactly the local speed of sound, so no rarefaction can catch and weaken it. The CJ state is found from the conservation (Rankine-Hugoniot) relations plus the tangency condition:

ρ₁ D = ρ₂ (D − u₂) (mass)
P₁ + ρ₁ D² = P₂ + ρ₂ (D − u₂)² (momentum)
h₁ + ½D² + q = h₂ + ½(D − u₂)² (energy, with heat release q)

For a stoichiometric hydrogen-air mix, D_CJ ≈ 1,970 m/s and the CJ pressure is ≈ 15.6 atm from 1 atm; ethylene-air gives ≈ 1,820 m/s. Thermodynamically, a PDE approximates the Humphrey cycle — Brayton but with constant-volume heat addition instead of constant-pressure. Adding heat at constant volume drives up pressure (ΔP ∝ T rise), and higher pressure at the point of heat release means more work is recoverable. That is the ~5–15% ideal thermal-efficiency edge over an equivalent Brayton engine.

One cycle, step by step

A single PDE tube runs a repeating four-part cycle, each phase measured in milliseconds:

• Fill — a fresh fuel-air (or fuel-oxygen) charge is injected to fill the tube, e.g. a 1 m tube in a few ms.
• Initiate & DDT — a spark or hot jet ignites a deflagration, which must accelerate and transition into a detonation (see next section) before it reaches the open end.
• Detonate & blow-down — the CJ wave sweeps to the open exit at ~2,000 m/s (crossing a 1 m tube in ~0.5 ms), then a rarefaction fan and the high-pressure products expand out the back, delivering the thrust impulse.
• Purge — a short slug of air scavenges hot residual gas so the next fresh charge does not pre-ignite.

Repeat this at 20–100 Hz and the string of impulses averages into usable thrust. A useful figure of merit is specific impulse per cycle: a partially-filled, air-breathing PDE tube can reach I_sp ≈ 1,000–4,000 s on hydrocarbon fuel, competitive with turbojets, without any turbomachinery.

The hard part: getting to detonation in a short tube (DDT)

You cannot simply spark a fuel-air mix and get a detonation — a spark makes a deflagration. Converting it is deflagration-to-detonation transition (DDT), and it is the central engineering problem of PDEs. Left alone, DDT can require several meters of run-up, far too long for a flightweight tube. Engineers force it with turbulence-generating obstacles: the classic device is a Shchelkin spiral, a helical coil inserted in the tube whose blockage ratio (typically 0.3–0.5) trips the flame into intense turbulence, folding and accelerating it until a local explosion births the detonation. A good spiral cuts DDT length by 50–57%, bringing it to tens of centimeters.

There is a direct trade: obstacles that promote DDT also throttle the fill flow and cause pressure loss during blow-down, cutting cycle frequency and specific impulse. In lab tests, a smooth 'clean' tube might only reach ~30% of CJ pressure (an incomplete transition), while the same tube with a spiral sustains full CJ spikes. Alternatives include machined orifice plates and pre-detonators (a small, easily-detonated oxygen-fuel initiator that jets into the main tube).

Real hardware, real numbers

The landmark demonstration was the Long-EZ 'Borealis', flown at Mojave on 31 January 2008 by the Air Force Research Laboratory and Innovative Scientific Solutions, Inc. — the first flight of an aircraft under pulse-detonation power. Its engine used four tubes each detonating at 20 Hz (80 Hz total across the tubes), producing up to ≈ 890 N (200 lbf) of thrust. It ran under its own power for about 10 seconds at ~30 m altitude (a rocket assist handled liftoff). The airframe is now in the National Museum of the U.S. Air Force.

PDE concepts have been studied for afterburner replacement, cruise-missile propulsion, and as combined-cycle stages toward hypersonics. In practice the field has largely pivoted to the rotating detonation engine (RDE) — a related device where a detonation wave runs continuously around an annulus at ~1–10 kHz — because it delivers the same near-constant-volume gain with steadier thrust and no valves. GE, Aerojet Rocketdyne, and multiple national labs have run RDE test articles; the pulsed variant remains mostly a research and demonstrator platform.

Trade-offs, failure modes, and a common misconception

Misconception: a PDE is 'just a supercharged pulsejet.' It is not. A V-1 pulsejet is a deflagration device — subsonic flame, mild pressure oscillation, low efficiency. A PDE runs true supersonic detonations and follows the Humphrey (constant-volume) cycle; the physics and the efficiency ceiling are different.

Key limitations engineers wrestle with:

• Noise and vibration. Near the tube exit, detonation pressure pulses reach roughly 170–190 dB (near the ~194 dB physical ceiling for undistorted sound in air); fatigue and acoustics are severe design constraints — Borealis had to prove the airframe survived them.
• Valving and fill. Mechanical valves at 80 Hz across many tubes are wear-limited; getting a clean, un-mixed fresh charge each cycle without residual hot-gas pre-ignition is delicate.
• Unsteady, poorly-matched nozzle. A single fixed nozzle cannot be optimal across the fill-detonate-purge cycle, so real specific impulse falls short of the ideal Humphrey value.
• DDT reliability. A misfired or late transition wastes a whole cycle's fuel and can quench the wave. Multi-tube staggering smooths thrust but multiplies the valve and initiation hardware.

The subtle pitfall: designers who tune a tube for maximum DDT robustness (dense obstacles, rich mixture) often destroy the very efficiency advantage that justified detonation in the first place — the win only appears when the wave reaches full CJ strength and the flow losses stay small.

Pulse detonation vs. conventional constant-pressure (Brayton) combustion propulsion

Attribute	Pulse Detonation Engine	Gas Turbine / Ramjet (Brayton)
Combustion mode	Detonation (supersonic, shock-coupled)	Deflagration (subsonic, ~1–100 m/s flame)
Thermodynamic cycle	Humphrey (near constant-volume)	Brayton (constant-pressure)
Pressure across combustor	Rises 15–20× (self-compressing)	Falls 3–8% (loss)
Moving compressor needed?	No — the wave compresses the charge	Yes (turbojet) / ram only (ramjet)
Thrust character	Pulsed at 20–100 Hz, high peak	Steady, continuous
Ideal efficiency gain	~5–15% higher than Brayton	Baseline
Maturity	Flight-demoed 2008; not fielded	Mature, in service for decades