20 kN Dual-Regen Cooling Aerospike Engine
Aerospike rocket engines are of serious interest, as they have higher nozzle efficiency in a wide range of altitudes. A traditional bell nozzle has a fixed expansion ratio, so the engineers have to optimise it for a specific part of the ascent profile. and the performance losses outside its optimum operating range follow. An aerospike engine works differently. When the vehicle accelerates, the ambient pressure forms the exhaust plume into a self-adapting virtual nozzle wall. In principle this can raise the overall ascent efficiency and thus increase mission performance.
The biggest challenge has always been thermal management. More surface area of extreme heat is exposed to the Aerospike engine than to the conventional bell nozzle. The spike or plug is also in one of the hardest flow environments in the propulsion system. Cooling is not a secondary design concern but one of the central engineering challenges. Modern aerospike demonstrators are significant because they are more than a revisiting of an old concept. They use advanced manufacturing, complex internal cooling channels and improved propellant plumbing to tackle heat loads that previous programmes could not tackle cost-effectively.
Aerospike Basics
A plug nozzle is an aerospike that channels hot gas along the outside of a centrebody. At low altitude, the atmospheric pressure squeezes the plume more tightly against the spike. But as the pressure decreases with altitude, the exhaust expands further, so the expansion ratio effectively increases during climb. The big plus is that altitude compensation.
This is important from a defense and security perspective, as effective margins create real options. They can buy more payload, more manoeuvre margin or more fuel for range.” In addition, greater efficiency may allow for a reduction in vehicle size for a given mission, which may simplify logistics and basing. Of course, none of that is certain. But the nozzle concept gives you a lever that bell nozzles can’t pull without a trade-off.
20 kN Test Scale Matters
A 20kN class engine is a good compromise. It is large enough to demonstrate real combustion and cooling behaviour. But it’s still small enough that you can iterate quickly and test often. This is perfect for proving whether a complex nozzle and cooling concept withstands repeated hot fires without nasty wear patterns. A high-profile example is the DemoP1 demonstrator, described as a 20 kN LOX/LNG aerospike with dual regenerative cooling and additive manufacturing at the heart of the build. Engineers also view it as an important milestone because it combines a complex aerospike design with modern manufacturing techniques, instead of serving as a one-off laboratory experiment.

Double-Regenerative Cooling
However, regenerative cooling is the default rocket trick—run the cryogenic propellant through channels in the walls of the engine before injection. The metal does not melt, and the propellant absorbs the heat. Meanwhile, the propellant comes out of the channels warmer and easier to handle at the injector. Regenerative cooling is a leading technology used for liquid rocket chambers and nozzles, often in combination with other techniques, such as film cooling, for control of peak heat flux (see NASA propulsion literature).
The cooling load is split in two different loops with an approach of double regeneration (double-circuit). Since with an aerospike you don’t have a single “wall”. You have the plug/spike surfaces. You have the outside structure. Each with their own heat maps and stress profiles. Two circuits allow designers to control the coolant flow to the areas of greatest heat flux, rather than over-cool one spot and risk hot spots in other areas.
Oxygen cools the internal plug circuit in the published DemoP1-like configuration, while methane/LNG cools the external structure. The divide is not skin deep. That’s a sensible decision. The spike is in the most punishing flow, so it needs its own cooling budget.

Where It Fits
The 20 kN Dual-Regen Cooling Aerospike Engine is optimised for efficiency at different altitudes. Small launchers can use it on upper stages to get more delta-v from compact tanks. Groups of 20 kN units can also provide power for responsive micro-launch vehicles, with modularity supporting production and spares. It could also be used by defense-relevant space tugs for multi-burn orbital manoeuvres for ISR replenishment and rapid payload repositioning. Finally, hypersonic testbeds could apply the cooling lessons and plug-nozzle flow control to high-heat propulsion demonstrators, where thermal durability drives program tempo.
The spike drives the entire design.
The exhaust temperatures of Aerospike can be very high. An aerospike test report shows the exhaust around the spike is about 3500°C, highlighting the engineering challenge of cooling it. Once you accept that thermal reality, the architecture falls into place. You want channels that follow the local heat flow. You want a predictable flow distribution. And you need stuff that can lose heat fast without falling apart. Otherwise you get an “efficient nozzle”, which is a short-lived heater.
Additive Manufacturing Makes It Possible
Aerospikes reward complex internal geometry. They punish simple plumbing. Cooling channels can be manufactured by traditional fabrication, but the more complex the shape, the slower and more expensive the process becomes. Additive manufacturing changes that equation.” It provides the capability to fabricate channel networks on curved surfaces and with cross-section variations where the heat flux is not uniform. It can also reduce part count, which often reduces leak paths and assembly risk.
This advantage is the reason for the importance of the LOX/LNG dual-regenerative demonstrator, even for its thrust class. “It’s not just an aerospike. It’s a manufacturing and thermal management argument. If you can print and qualify the hot section reliably, aerospikes go from an alluring to a producible concept.”

Performance Gains Bring Trade-Offs
Aerospikes can improve overall ascent efficiency because they compensate for altitude. However, they do not give you free performance. You still pay in:
- Cooling is complex: more hot surface area requires more careful design.
- Weight of structure. More material, more plumbing, generally a larger wall area.
- Integration and base heating are problems, particularly if the spike is shortened to save weight and length.
So the real question is not “does the nozzle concept work?” It does. The real question is, “Can you run it over and over, look at it easily, and put it together at a cost that makes sense?” That is the question every reusable propulsion programme must answer.
Why It’s Pivotal for Defense
Aerospike engines are often dismissed as niche ‘space nerd’ technology, but the relevance to defence is far from academic. Efficient propulsion enables rapid launch, quick ISR satellite replenishment, and more resilient space architectures. This is important not only for launch vehicles. “High-temperature materials, cooling design, and thermal management are also areas that could be advanced by similar research for hypersonic research vehicles and other extreme-flight systems. This technical overlap helps explain why companies like POLARIS talk about aerospikes in connection with the development of reusable spaceplanes. Aerospikes have a long history of testing, and the renewed interest is supported by real flight data, including work led by NASA on the performance of aerospike nozzles. That is important because measured data reduces engineering risk far more than concept slides or theoretical claims.

2026 Watchpoints That Matter
If you want to judge whether a Double Regen Cooling 20 kN Aerospike Engine is heading towards operational relevance, watch for four concrete signals:
- The engine would need to show repeatable hot-fire campaigns with consistent performance and no progressive hot-spot damage.
- “Post-test inspections give important indications of results, including initiation of cracks near channel corners and manifolds.
- Transient stability at throttle and start, as aerospikes may have different behaviour during ramp-up and shutdown.
- Manufacturing yield (number of printed hot sections that pass NDT and pressure checks without rework)
Meanwhile, teams can show steady progress in those areas, and aerospikes gain credibility fast. If they cannot, the concept returns to the drawer—again.

Aerospike vs Bell-Nozzle Engines
| Spec / Parameter | Aerospike nozzle rocket engine | Conventional bell-nozzle rocket engine |
|---|---|---|
| Nozzle geometry | Plug/centrebody (annular) or linear ramp | Bell-shaped nozzle |
| Altitude compensation | Yes (plume adapts with ambient pressure) | No (fixed expansion ratio) |
| Common propellants (examples) | LOX/CH₄ (methalox), LOX/LNG, LOX/RP-1, LOX/LH₂, NTO/MMH (less common for modern aerospike demos) | LOX/RP-1, LOX/CH₄, LOX/LH₂, NTO/MMH, N₂O₄/UDMH, HTP/kerosene (varies by programme) |
| Propellant choice impact | Cooling and materials can drive choices; dual-circuit regen often pairs well with LOX + methane/LNG because both can be used as coolants | Very flexible; many mature designs across all major propellant families |
| Best efficiency region | Strong across a wide altitude range | Strong in a narrow design altitude band |
| Sea-level behaviour | Often avoids severe over-expansion issues | Over-expansion/flow separation can occur if vacuum-optimised |
| Vacuum performance | Often comparable to a good vacuum bell, not automatically higher | Excellent when vacuum-optimised |
| Hot surface area exposed | Higher (spike + outer flow surfaces) | Lower for similar thrust |
| Cooling demand | Harder (spike hot spots; complex heat map) | Simpler (more uniform nozzle wall cooling) |
| Typical cooling approach | Aggressive regen; sometimes dual-circuit regen + film cooling | Regen + film cooling common; fewer circuits needed |
| Manufacturing complexity | High; complex channels/geometry (AM helps a lot) | Lower, highly mature tooling and QA |
| Mass & packaging | Can be heavier/complex, but can be package differently | Often lighter and simpler for equivalent thrust |
| Thrust vector control | Can be trickier (methods vary by design) | Usually straightforward (gimbal) |
| Base heating/plume interaction | This process can be more challenging, particularly when truncation occurs. | While aerospikes are more predictable, they still require effective management of base heating. |
Conclusion
Aerospikes didn’t fail in theory. They failed in practice. Their theoretical performance advantages were always obvious, but the real trick was to survive the thermal environment. At the 20 kN demonstrator level, advanced manufacturing and dual-circuit regenerative cooling give designers better tools to handle that heat today.
So the 20 kN double-regenerative cooled aerospike engine is not just a propulsion experiment. It is a stress test for modern thermal engineering. How can engineers cool the spike? Can they stabilise flow in a channel under extreme conditions? Can they reliably build the hot section again and again? If so, then aerospikes begin to move from curiosity to practical capability.
References
- https://www.metal-am.com/articles/making-the-unmakeablehow-3d-printing-is-bringing-the-aerospike-rocket-engine-to-life/
- https://www.sciencedirect.com/science/article/pii/S1359431125017764
- https://polaris-raumflugzeuge.de/Technology/Aerospike-Engines
- https://www.nasa.gov/news-release/nasa-dryden-flight-research-center-news-room-news-releases-aerospike-engine-flight-test-successful/
- https://www.hse.gov.uk/comah/sragtech/techmeasndt.htm
- https://apps.dtic.mil/sti/tr/pdf/ADA609649.pdf

