Scramjet Inlet

The job of a scramjet inlet

The job is brutally specific. Take air entering at Mach 8 and 220 K (the freestream conditions of a scramjet at 80,000 ft altitude), deliver it to the combustor at Mach 2.5–3 and roughly 1500 K, and lose as little total pressure as possible along the way. There is no compressor. There is no rotating part anywhere in the engine. The compression is done entirely by the shape of the duct and the laws of compressible flow.

A turbojet's compressor at sea level might raise stagnation pressure by a factor of 40 across a dozen rotating stages. A scramjet inlet at Mach 8 raises static pressure by a factor of 30 across three or four stationary ramps, in a length of perhaps two metres, with the only moving thing being the air itself. The trade is that the air arrives at the combustor still moving faster than sound, and combustion has to happen on a stopwatch.

Why not finish with a normal shock?

A pure ramjet decelerates flow to subsonic at the combustor by ending its inlet with a terminal normal shock. That works up to about Mach 5. Beyond Mach 5, the normal-shock total-pressure recovery (Rayleigh-Pitot formula) is a death sentence for the cycle:

Freestream Mach	Normal shock total-pressure recovery
2.0	72 %
3.0	33 %
5.0	6 %
6.0	3 %
8.0	1.5 %
10.0	0.8 %

By Mach 6 there is essentially no pressure ratio left to expand through the nozzle. The engine cycle is dead. The fix — the only fix — is to never go normal. The scramjet inlet uses only oblique shocks, and each oblique shock has a much higher recovery because only the velocity component normal to the shock matters thermodynamically. A 10° oblique shock at Mach 8 has a recovery around 90 percent; stacking three of them still leaves you with ~70 percent total pressure recovery, against 1.5 percent for a single normal shock. The price is that the air never goes subsonic.

Station by station, Mach 8 to combustor

A representative scramjet inlet at design Mach 8 follows roughly this path:

Station	What it is	Mach	Static T (K)	Static p (relative)
0	Freestream	8.0	220	1.0×
1	After ramp 1 (forebody, 6°)	6.4	360	2.4×
2	After ramp 2 (8°)	5.1	620	6.5×
3	After ramp 3 (10°)	3.9	1050	17×
4	Cowl lip + reflection	3.2	1300	26×
5	Isolator exit / combustor face	2.8	1500	30×

The flow has been decelerated by a factor of three, heated by a factor of seven (compression heating, not combustion), and squeezed by a factor of thirty — all without bringing it subsonic, and all with two stationary surfaces and a cowl. The combustor face sees air at Mach 2.8, hot enough that hydrogen will spontaneously ignite without a separate ignition source.

Shock-on-lip and forebody compression

The geometry is tuned so that at the design Mach number every oblique shock falls exactly on the cowl lip. No air spills outside the inlet (no spillage drag); no shock is captured at a position that produces a non-uniform combustor-face profile. This is called shock-on-lip design. Off the design Mach, the shocks either fall short of the lip (spillage) or impinge inside the duct (loss). The penalty for being off-design at Mach 6 when you were designed for Mach 8 can be a 20 percent capture-efficiency hit.

In a vehicle-integrated scramjet, the first compression ramp is not on the engine — it's the airframe's underside. The vehicle is shaped to ride on its own bow shock (a waverider), and the lower forebody acts as a single, long compression surface that captures air into the engine intake. X-51A used exactly this scheme: the cylindrical vehicle had a wedge underside that started the compression, the engine pod hanging beneath finished it. This is also why a scramjet cannot be redesigned in isolation from its airframe — they are aerodynamically a single object.

The isolator: a duct that absorbs pressure waves

Between the inlet's last shock and the combustor's first fuel injector sits a constant-area duct called the isolator. Its job is to absorb the back-pressure pulse from combustion. When fuel lights, the chamber pressure spikes upstream; without the isolator, that pulse would walk all the way back into the inlet and trigger unstart. The isolator typically contains a stable shock train — alternating oblique compression and expansion regions — that holds the pressure rise within its length.

Isolator length scales with how strong the heat release is. In ramjet mode (lower flight Mach, large heat release relative to dynamic pressure) the isolator may need to be eight to twelve duct diameters long to hold the shock train. In pure scramjet mode (higher flight Mach, smaller fractional heat release) the same hardware may run with a much shorter effective shock train — sometimes just a few diameters. A dual-mode engine sizes the isolator for the ramjet case, where it matters most.

Unstart: the only failure mode that matters

Unstart is the death of a hypersonic inlet. In normal operation, the inlet contains a swallowed shock train sitting near the throat: a series of oblique compressions confined inside the duct. If anything raises back-pressure beyond the isolator's capacity — too much heat release, a fuel-mixing instability, vehicle yaw that misaligns the inlet, throat icing — the shock train walks forward. Once it reaches the cowl lip, it pops out of the duct, becomes a detached bow shock standing in front of the inlet, and the entire engine inhales perhaps half the mass flow it was designed for.

Consequences arrive in sequence over milliseconds. Thrust collapses to near zero. The vehicle decelerates, but asymmetrically (one engine of a multi-engine vehicle may unstart while the others stay started), producing yaw moments large enough to break the airframe. Structural loads from the detached shock can exceed design limits by factors of two or three. SR-71 pilots called it the "inlet hammer" and described it as being kicked in the back by a horse — and the SR-71 was only at Mach 3.

Modern scramjets monitor pressure at multiple stations in the isolator with hundreds of microsecond response time. Onboard control logic detects shock-train migration before it reaches the lip and responds by cutting fuel flow to lower back-pressure, in effect re-starting the inlet without losing the vehicle.

Worked example: total-pressure budget at Mach 8

Consider a scramjet inlet designed for Mach 8 freestream, with three external compression ramps of 6°, 8°, and 10°, ending with a cowl lip that catches the third shock and reflects it into the isolator.

Ramp     Pre-ramp M    Post-ramp M    p02/p01
1 (6°)        8.0            6.4         0.96
2 (8°)        6.4            5.1         0.94
3 (10°)       5.1            3.9         0.89
Cowl ref.     3.9            3.2         0.92
Isolator      3.2            2.8         0.95
Cumulative                                ≈ 0.70

Cumulative total-pressure recovery: 70 percent. For comparison, a single normal shock at Mach 8 would deliver about 1.5 percent. The oblique-shock train recovers roughly 47 times more total pressure than the equivalent single normal shock — which is why scramjets, and only scramjets, work above Mach 6.

Real engines, real flights

• NASA X-43A (Hyper-X), 2004. The first scramjet to fly under its own power. A 3.7-metre, 1,300-kg hydrogen-burning waverider, dropped from a modified Pegasus booster launched from a NASA B-52B at 40,000 ft. On its third flight (16 November 2004) it accelerated to Mach 9.6 — about 11,200 km/h, still the speed record for an air-breathing aircraft — and sustained powered flight for roughly 10 seconds before the engine shut down and the vehicle gilded to splashdown in the Pacific. Cost of the Hyper-X program: $230 million for three flight tests.
• USAF X-51A Waverider, 2010 – 2013. A larger (4.3 m, ~1,800 kg) hydrocarbon-fuelled scramjet using JP-7 — the same fuel that fed the SR-71. Four flights between 2010 and 2013. Flight 4 (1 May 2013) sustained Mach 5.1 for 240 seconds — the longest scramjet-powered flight ever — covering 426 km before splashing into the Pacific by design. The X-51A demonstrated that hydrocarbon scramjets work for minutes, not just seconds, which is the operational regime a hypersonic weapon or cruiser requires.
• HIFiRE 7 (Australia / US), 2017. A sounding-rocket-launched scramjet flight that reached Mach 8.2 in 2017 under joint Australian DST Group / US Air Force funding. Produced much of the open-literature data on hydrocarbon scramjet combustion at flight conditions.
• 3M22 Zircon (Russia, operational 2023). A reported scramjet-cruise anti-ship and land-attack missile with a stated Mach 8–9 capability. Operational details are classified; Russian state media has shown launches but not flight telemetry.
• HACM (US Air Force, in development). The Hypersonic Attack Cruise Missile, a planned air-launched scramjet missile drawing on X-51A heritage. First flight planned for the late 2020s.

Inlet types compared

Inlet	Mach range	Final shock	Combustor flow	Total-p recovery at design
Subsonic (turbojet)	0 – 0.8	None	Subsonic	~ 99 %
Pitot (low-supersonic)	1.0 – 1.5	Normal at lip	Subsonic	~ 95 %
Single-ramp supersonic	1.5 – 2.5	Normal at throat	Subsonic	80 – 90 %
Multi-ramp ramjet	2 – 5	Normal at throat	Subsonic	40 – 70 %
Dual-mode (X-51A)	4 – 7	Thermal throat / none	Sub or supersonic	50 – 70 %
Pure scramjet (X-43A)	5 – 15+	None — all oblique	Supersonic (M 2.5 – 3)	30 – 60 %

The last row is the scramjet's regime, and it is the only known inlet type that produces net thrust above Mach 6 without rocket-style onboard oxidizer.

Design knobs and their consequences

• Ramp angles. More ramps with smaller angles → higher total-pressure recovery but longer, heavier engine. Most flight vehicles converge on 3–4 external ramps plus a cowl-lip reflection.
• Capture area ratio. The inlet capture area divided by combustor face area. Typically 5–10 for scramjets — the inlet swallows much more frontal area than it delivers, which is how it produces compression without rotating parts.
• Cowl angle. Trade between drag (steep cowl → more drag) and capture efficiency (shallow cowl → spillage at off-design Mach).
• Isolator length. Longer → more margin against unstart but more wetted-area friction and structural weight. A representative isolator runs 8–12 duct diameters in a dual-mode engine.
• Forebody integration. Whether the vehicle's underside acts as the first compression ramp. Waveriders win on capture-area efficiency but couple airframe and engine design.

The thermal environment

At Mach 8, the stagnation temperature of the air — what it would reach if brought adiabatically to rest — is around 2700 K. The inlet ramps see static temperatures from 220 K at the leading edge up to roughly 1500 K at the cowl lip; the actual surface temperature, after accounting for skin friction and recovery factor, can exceed 1800 K for sustained flight. Titanium melts at 1940 K. Inconel softens at 1300 K. No metal works for a hypersonic inlet at design Mach without active cooling.

X-43A and X-51A used carbon-carbon composites and ultra-high-temperature ceramics (UHTCs — zirconium diboride, hafnium carbide) for the leading edges and most-loaded ramps. For longer-duration vehicles, the fuel is routed through cooling channels in the wall before reaching the injectors, a scheme called regenerative cooling, identical in principle to the cryogenic-fuel cooling of a rocket nozzle. The fuel becomes both propellant and coolant — the only chemistry that closes the energy balance on a sustained hypersonic flight.

Common misconceptions

• The combustor goes subsonic. Not in a true scramjet. The whole flow path is supersonic from inlet to nozzle exit. The "S" in scramjet is the supersonic combustor.
• An inlet is just a hole. The inlet does all of the compression work. It is the most aerodynamically demanding subsystem on a hypersonic vehicle.
• You can run a scramjet from a standstill. A scramjet at zero velocity has no compression and no combustion. Every flight test has used a rocket booster to get to scramjet light-off Mach (typically Mach 4+).
• Fixed geometry can't span Mach 4 – 10. Real fixed-geometry scramjet inlets do span this range, by accepting capture-efficiency loss away from the design Mach. Variable-geometry inlets exist (the SR-71's translating spike) but add mechanical complexity that scramjets are designed to avoid.
• Unstart only happens at the wrong Mach. Unstart is back-pressure driven. It can happen at design Mach if combustion runs too hot, or if the vehicle yaws, or if the fuel-air mixture goes off.
• Hypersonic propulsion is exotic future tech. Scramjet inlets have been flying since 2004. The engineering problem is duration and reusability, not feasibility.