You can't hold a detonation still. But you can start one, let it run down a tube, and start another — over and over, like a machine gun. On 31 January 2008, a small aircraft flew over the Mojave Desert for ten seconds pushed by exactly that. Why not just fire a tube of detonations, again and again?
A cycle in four strokes
A Pulse detonation engineA tube that is filled, detonated, emptied and refilled many times a second. The precursor of the rotating detonation engine. is a tube with a valve at one end and an open nozzle at the other. Each cycle has four strokes, a little like a piston engine's:
- Fill. The valve opens and fresh fuel–oxidizer mixture flows into the tube.
- Detonate. The valve closes; an initiator (often a small pre-detonator, or a spark with a DDT section) starts a detonation at the closed end. It runs down the tube at the CJ speed — about 2 km/s — in half a millisecond.
- Blow down. The high-pressure products rush out of the open end, while an expansion wave travels back to the closed end and lowers the pressure there. This BlowdownThe part of a pulse detonation cycle in which the high-pressure products rush out of the open end while an expansion wave runs back up the tube. takes several times longer than the detonation.
- Purge. A slug of cool air separates the hot products from the next charge, so the fresh mixture is not ignited early by contact with hot gas.
The thrust comes from the closed end. Behind the Taylor expansion of Lesson 4, the products near the thrust wall sit at a plateau of nearly 40% (about three-eighths) of the CJ pressure, and they push on the wall until the blowdown's expansion wave arrives. Every cycle delivers one push.
Impulse per bang, times bangs per second
A pulse detonation engine's average thrust is simply its impulse per cycle times its firing frequency. Wintenberger, Shepherd and colleagues at Caltech showed that the impulse of a fully filled tube is well captured by a single formula: , where is the plateau pressure and .
Why fill only part of the tube?
The partial-fill effect is one of the PDE's surprises. Filling only part of the tube and leaving the rest full of inert gas lowers the impulse per cycle, but raises the impulse per unit of fuel: the inert gas acts as a tamper, which the detonation products push on before they escape, so more of their energy goes into momentum. The penalty is less thrust per tube. Try it in the cycle simulation above.
Why PDEs never quite made it
PDEs were studied intensively from the 1990s to the 2010s, by the US Air Force, NASA, the Navy, General Electric and Pratt & Whitney among others. In 2008 the Air Force Research Laboratory flew a modified Long-EZ aircraft powered by a four-tube PDE. But the engine's problems were stubborn:
- Initiation every cycle. Each cycle needs DDT, which takes length (Lesson 6), or a pre-detonator, which takes oxygen and complexity.
- Valves must open and close tens to hundreds of times a second against detonation pressures and temperatures.
- Low frequency and pulsing thrust, with the hardware vibrating at the firing rate.
- Most of the cycle is spent not detonating — filling, purging and blowing down.
What if the detonation never had to stop? What if it could keep running into fresh mixture, continuously, while the mixture was being replaced behind it? That idea — first realised by Voitsekhovskii in 1959 — is the rotating detonation engine.
- A PDE cycles through fill, detonate, blow down and purge; thrust comes from the plateau pressure on the closed end.
- Impulse per cycle ≈ ; thrust = impulse × frequency.
- Filling dominates the cycle, so PDEs fire at tens to a hundred hertz per tube.
- Partial filling raises impulse per unit fuel at the cost of impulse per cycle.
- Valves, repeated initiation and low frequency held PDEs back — motivating a detonation that never stops.
Next: bend the tube into a ring.