The Reaction Engines SABRE and the Precooled Engine That Could Fly to Orbit

The most expensive part of reaching orbit has always been the oxidizer you have to carry with you. Reaction Engines Limited spent three decades developing an engine that changes that equation: the Synergetic Air-Breathing Rocket Engine, or SABRE - a propulsion system that breathes atmospheric air like a jet up to Mach 5, then switches to onboard oxidizer to continue to orbital velocity. In April 2019, an independent test validated the engine’s critical enabling technology. In mid-2023, the company entered administration. The physics, however, remain intact.

Why Carrying Your Own Oxygen Is Such a Problem

To reach low Earth orbit, a vehicle must accelerate to approximately 17,500 mph - Mach 23. Conventional rockets achieve this by carrying both fuel and oxidizer, since there is no atmospheric oxygen available in space. The oxidizer, typically liquid oxygen, is dense, heavy, and must be refrigerated. On a typical orbital rocket, propellant - fuel and oxidizer combined - accounts for 80 to 90 percent of total vehicle mass at liftoff. Useful payload is a small fraction of what you started with.

For the early portion of any orbital trajectory, however, the vehicle is still inside the atmosphere. The air is thin above 60,000 feet, but oxygen is present. An engine capable of using that atmospheric oxygen during ascent - and switching to stored oxidizer only when the atmosphere can no longer support combustion - would carry significantly less propellant from the ground and deliver a much more favorable mass fraction to orbit.

That is the premise behind SABRE.

How the Engine Actually Works

In jet mode, operating below roughly Mach 5 and 25 kilometers altitude, SABRE compresses incoming atmospheric air, mixes it with liquid hydrogen fuel, and combusts the mixture to produce thrust. No liquid oxygen tanks are required for this phase. Above those thresholds, where air density drops too low and inlet conditions become too extreme, an inlet closes and the liquid oxygen supply opens. The engine transitions to a conventional rocket cycle and continues accelerating to orbital velocity.

Reaction Engines designed a vehicle concept around two SABRE engines called Skylon: a slender lifting body approximately 80 meters long, with engine pods mounted at mid-fuselage. The design called for runway takeoff, acceleration to Mach 5 at altitude in jet mode, transition to rocket mode, orbital payload delivery, atmospheric reentry, and runway landing. Single-stage-to-orbit. No expendable hardware.

The Thermodynamic Wall That Blocked This for Decades

Making a jet engine function at Mach 5 inlet conditions requires solving a specific, hard problem. At that velocity, the stagnation temperature - the temperature incoming air reaches as it decelerates and compresses at the inlet - reaches approximately 1,000°C (roughly 1,800°F). That exceeds the melting point of aluminum and degrades many steel alloys. Conventional turbomachinery cannot operate in that environment.

This is why sustained air-breathing propulsion above Mach 3 to 3.5 has historically been so difficult. Concorde’s Rolls-Royce Olympus turbojets operated at around Mach 2. The SR-71 Blackbird’s Pratt & Whitney J58 engines used a hybrid turbojet-ramjet architecture to reach Mach 3.2, but required exotic materials and extraordinary engineering. Mach 5 with turbomachinery appeared to be a wall.

The Precooler: What Makes SABRE Different

Reaction Engines Limited was founded in 1989 by three British aerospace engineers: Alan Bond, Richard Varvill, and John Scott-Wilson. Bond had previously worked on an earlier concept called HOTOL (Horizontal Take-Off and Landing), which pursued similar goals but was blocked by two factors: a British government security classification on key engine design details, and an unsolved thermal management problem. The founders of Reaction Engines began again from scratch.

Their solution was the SABRE precooler - a heat exchanger that takes incoming air at ~1,000°C and cools it to approximately -150°C before it reaches the turbomachinery. The temperature drop of over 1,100°C occurs in a fraction of a second, inside a compact heat exchanger light enough to fly and durable enough to survive repeated thermal cycling across that entire range.

The cooling medium is supercritical helium flowing through a network of tubes measured in fractions of a millimeter in diameter. Manufacturing those heat exchangers at the required tolerances, at production scale, is a problem in metrology, materials science, and manufacturing process simultaneously.

The Icing Problem Inside the Icing Problem

Cooling atmospheric air from 1,000°C to -150°C nearly instantaneously creates a secondary problem. Air contains water vapor. At that rate of temperature drop, water vapor wants to nucleate into ice crystals - which block heat exchanger channels, damage surfaces, and destroy the component.

This was the thermodynamic trap that blocked earlier attempts at the concept. Solving the inlet temperature problem opened the icing problem. It appeared to be a dead end.

Reaction Engines’ answer was to inject small quantities of hydrocarbon compounds into the incoming airstream upstream of the precooler. The injection prevents water vapor from nucleating into ice crystals. The moisture passes through without freezing.

The 2019 Test: Independent Validation

In April 2019, Reaction Engines operated a precooler test article at their facility in England under conditions representing Mach 3.3 inlet temperatures. Incoming air was heated to approximately 420°C to simulate hypersonic flight at that velocity. The precooler cooled it. The anti-icing system prevented frost formation. The system performed as modeled.

Independent technical observers from the European Space Agency and the U.S. Air Force Research Laboratory were present - institutions with personnel there specifically to identify problems, not validate assumptions. The test held up.

This was the first independently validated demonstration that the SABRE precooler concept was physically real. Not a simulation or a theoretical paper. An actual piece of hardware doing something the aerospace engineering community had been debating for thirty years.

Investment and Defense Interest

The 2019 test did not occur in isolation. BAE Systems had taken a 20 percent equity stake in Reaction Engines in 2015. Boeing’s investment arm put capital in during 2018. The UK Space Agency had been a consistent funding source through research contracts. DARPA had been paying attention for years, and the U.S. Air Force Research Laboratory formalized a cooperative research relationship with the company.

When a technology can sustain air-breathing propulsion at hypersonic speeds, the applications extend well beyond a spaceplane. High-speed strike platforms, long-range reconnaissance, hypersonic testbeds - the defense interest was not casual.

Administration and What It Means

In mid-2023, Reaction Engines Limited entered administration - the British legal equivalent of bankruptcy proceedings. The company ceased operations in its existing form.

The company failed. The technology did not.

The intellectual property entered the administrative process and has been subject to acquisition interest. Engineers who spent careers on SABRE have continued working in related areas. The precooler concept, the icing solution, the test data, the analytical framework for precooled hybrid propulsion - all of it exists in published technical literature. In AIAA conference archives. In Acta Astronautica. In the European Space Journal. In Reaction Engines’ own documentation, much of which remains publicly available through ESA and AIAA libraries.

Where Precooled Propulsion Stands Now

The concept has spread beyond one company. JAXA (Japan Aerospace Exploration Agency) has been developing a pre-cooled turbojet applying similar thermodynamic principles to high-speed propulsion. Research groups in China have published on precooled engine architectures. The European Space Agency maintains active interest in air-breathing access-to-orbit technologies. DARPA continues funding hypersonic propulsion research grounded in the same underlying physics.

Skylon may never be built. The SABRE engine as Reaction Engines designed it may never fly. But the gap that remains is capital and time - not unsolved physics. The hardest engineering problem at the center of the concept, the one that stopped every earlier attempt, has been solved and independently verified.

Why This Matters for Pilots

Every piston and turbine aircraft operates by using the atmosphere as a working fluid. The engine breathes air. The wings generate lift from air. The atmosphere is not an obstacle - it is the medium you operate in.

SABRE was an attempt to extend that relationship to the edge of space. To keep using the atmosphere as a resource for as long as it cooperates, and abandon it only when required. To make the transition from pilot to astronaut a gradual one, not a categorical leap.

That idea has not been disproven. The next program that advances precooled hybrid propulsion will build on what Reaction Engines published. That is how engineering knowledge propagates even when the organization behind it does not survive.

Key Takeaways

• SABRE is a hybrid engine that breathes atmospheric air as a jet up to Mach 5, then switches to onboard liquid oxygen for the rocket phase to orbit - eliminating the need to carry oxidizer for the lower atmosphere portion of flight.
• The core barrier to this concept was inlet stagnation temperature at Mach 5 (~1,000°C); Reaction Engines’ precooler solves this by dropping the air to -150°C in a fraction of a second using supercritical helium.
• A secondary icing problem - water vapor nucleating inside the precooler - was solved through upstream hydrocarbon injection, a breakthrough now part of the public technical record.
• In April 2019, the precooler was independently validated by ESA and USAFRL observers in a live hardware test - the first such demonstration after 30 years of debate.
• Reaction Engines entered administration in mid-2023, but the validated physics, test data, and engineering literature remain in the public domain and are actively informing parallel programs at JAXA, in Europe, and in U.S. defense research.