Five seconds past Mach 1on air alone.
A 25 kg aircraft, no rockets, level flight through the transonic wall. This is what the literature says actually gets you there, and what it says does not.
The short version
The thread opened on a turbojet-boosted ramjet that hands off below Mach 1. The evidence pulled it somewhere plainer: a bought micro turbojet, a short augmentation burst, and an airframe shaped around transonic drag. The hard part is not the engine. It is keeping thrust margin from Mach 0.8 to 1.1.
The rules are
the design brief.
A qualifying attempt is two flights by the same aircraft between sunrise and sunset of one day, watched by the judges. The prize goes to the first verified attempt. Every rule below removes a shortcut.
True airspeed above the local speed of sound, on each of the two flights. The supersonic legs must be flown on reciprocal headings, 180 degrees plus or minus 10.
55 lb max takeoff weight, including fuel and landing gear. Only fuel may be shed in flight.
Mach 0.8 to past Mach 1 in level or climbing flight. Altitude may not drop more than 10 m below the highest point since that phase began.
Controlled landing, wheels or belly, then fly again the same day. Refuel, recharge, minor repairs only. No new wings, fuselage, engine or flight computer.
Calibrated pitot-static plus total air temperature, sealed loggers, GPS, ground weather data, and the raw files. The margin covers measurement uncertainty.
Lift from aerodynamic surfaces. The team must build the majority of the airframe, in the spirit of FAA amateur-built rules (14 CFR 21.191(g)). Bought engines, servos and avionics are fine and do not count against that majority. Kits and target-drone derivatives are out.
Autopilot is allowed, but a remote pilot keeps command authority and an immediate abort on every flight.
Liquid fuel only, with gaseous hydrogen and nitromethane allowed as exceptions. Open to U.S. persons; minors need a parent or guardian co-signature. No venture capital, sponsors or grants; personal funds and no-equity crowdfunding are fine. Contest runs until it is awarded or December 31, 2030. This is the summary pasted into the thread; the binding text is at boomsupersonic.com/prize.
Two loopholes,
closed by email.
The question, Oct 15
- Can an external ground launcher release the aircraft below Mach 0.8, with everything after that done on air-breathing thrust? The ideas were a motorcycle engine or something like SpinLaunch.
- Can a pressurized CO2 cold-gas thruster handle attitude control during takeoff?
The aircraft must take off under its own power, so an external ground launcher is not permitted. ... Thrust from stored high pressure gas is prohibited, as is any non air-breathing thrust at any point in flight.
Boom Supersonic Prize Team reply, Oct 16, as quoted in the thread.
So the aircraft has to get itself off the ground and through Mach 0.8 to 1 on its own engine. Cold gas is out even for attitude control. That leaves the fixed-wing micro turbojet path from Query 3 as the strongest option.
Why the ramjet
idea broke.
The first answer proposed a turbojet that hands off to a ramjet around Mach 0.6 to 0.8. The papers put every real ramjet handoff far to the right of the contest window. Plot them on one axis and the problem is obvious.
Five ways to make thrust,
ranked by evidence.
Ranked for a Boom-scale aircraft, by how much small-scale testing backs each one, not by how exciting it sounds.
Turbojet + afterburner or duct burner
Most plausibleSmall-engine augmentation is directly studied. Variable nozzles and short-burst thrust gains fit a 5 second requirement. Apke 2012Shea 2024
Turbojet + ejector nozzle
Helper, not the answerCompact, no moving parts, but the gain falls off strongly with airspeed. Mostly a takeoff and low-speed benefit. Schmidt 2021
Turbojet + ramjet (TBCC)
Long shot, distinctiveMature as a concept, but the evidence centers on Mach 2.5+ transition, not a subsonic or transonic handoff. Liu 2024Lv 2022
Pulsejet / ram-pulsejet
Too risky for Mach 1Simple and light, with low efficiency and poor thrust density. No paper here shows a practical small supersonic aircraft on one. Piancastelli 2018Geng 2007
Air-breathing detonation
Unique but prematurePromising thrust in rigs and wind tunnels. Nothing yet closes the gap to repeatable small-aircraft supersonic flight. Frolov 2019Meng 2022Ivanov 2021
Inside the engine
you should buy.
A real, open KJ66 micro turbojet model. The thread describes the JetCat P140 class as representative of the sub-500 N, KJ66-derived architecture. Chachurski 2024 Each marker ties a part to what the research says about it.
The short-burst
augmentation stack.
Keep the turbine core and add thrust only where the drag rise needs it. These are the add-ons with real test or design work behind them, and what each one costs you.
A Colorado mini-turbojet project paired an afterburner with a variable-area convergent nozzle to run wet and dry efficiently. A related concept paired afterburning with nozzle control for a Mach 1.4 class vehicle. Apke 2012Walter 2012 Modular afterburners for JetCat-class engines are an active design topic. Shea 2024
Cost: thermal and pressure-loss management. The thread supports its plausibility more than a verified micro-engine gain number.
Reshaping the tail cone and using 2.5 degree struts reduced swirl while keeping diffuser pressure loss under 5%. Lee 2025
Cost: this is a subsystem result, not a flight result.
SFC dropped about 2.5 to 5.5% in the same tests. Silivestru 2025
Cost: benefit drops significantly with airspeed, so it helps takeoff and early climb, not the final push. Schmidt 2021
Gear-shaped high-g combustor work on micro turbojets keeps combustion performance at the same thrust while shortening the chamber. Huang 2024
Cost: integration maturity is still low.
Real, ground-tested hardware. Frolov 2019
Cost: integration and sensitivity to pressure losses.
Which turbine
to actually buy.
Published thrust for commercial engines that have test or modeling data behind them. The P300 class is the smallest with substantial experimental and CFD-validated support for serious transonic UAV work. Rhee 2020bLee 2023Lee 2024
Recommendation
A P300-class engine is the floor. A P400-class or PBS TJ100 / TJ150-class engine is more likely to keep transonic thrust margin if the airframe stays under 25 kg. Rhee 2020bCatană 2024Prisăcariu 2026
The reason is margin. The small-supersonic studies keep naming insufficient transonic thrust margin as the central risk. Mizobata 2016Dalman 2021Minato 2005 JetCat is the best documented brand in this set; PBS is described as especially mature for UAV integration. Prisăcariu 2026Catană 2024 Catalog thrust is not enough to size from: secondary air, lubrication fuel and off-design effects change thrust, SFC and temperatures, so pair test-stand data with a model. Aksoy 2026Vouros 2025Wu 2023Zhang 2025
The airframe matters
more than the engine.
The most repeated lesson across small supersonic test programs: getting supersonic depends on transonic drag reduction. Exotic combustion does not fix a draggy shape.
- 5 to 20%Drag reduction from area-rule shaping in Mach 0.8 to 1.2 analyses of Muroran's small supersonic testbed. Mizobata 2016
- 57 countsDrag cut at Mach 1.1 from a sharpened conical nose and a large-radius bottleneck. Mizobata 2024
- CancelledThe area-rule benefit on some configurations, once an intake was added. Design the inlet with the fuselage from the start. Mizobata 2024Mizobata 2025
- 61 to 66°Sweep of the cranked-arrow wing with a 6% diamond section used on the Muroran flight-test vehicles. Mizobata 2025Mizobata 2024
- 0.5 to 0.85%Supersonic drag penalty for enforcing low-speed stability in RANS-based optimization. Worth paying, because it has to take off and land. Seraj 2025
- 13.41 kgSmallest feasible supersonic UAV in one conceptual design framework. No feasible vehicles hit Mach 1.1 at 6 km. Aim for the mid-teens to low 20s. Dalman 2021
What that means for the shape
One slender fixed-wing airframe built only for this mission: small thin wing, high sweep, a waisted fuselage around the wing, and no VTOL requirement. Jacob 2021Mizobata 2005Mizobata 2014 The engine install is a primary design task, so inlet and exhaust nozzle are redesigned around the engine instead of bolted on. Compton 2021Hutchinson 2021
Control risk lives in the subsonic and transonic range. Directional stability degrades at large angles of attack and sideslip, elevon roll inputs can cause adverse yaw, and roll reversal can appear without coordinated rudder. Mizobata 2016Mizobata 2025 That is why groups fly subsonic prototypes first and check them against 6-DOF or point-mass predictions. García 2024Mizobata 2011aMizobata 2011b
For design tools, VLM stays useful for fast iteration while full CFD handles the harder supersonic features. Joshi 2023García 2024Jacob 2021
The detour:
a 590 km/h quadcopter.
The SQ-2 speed quad brief was the starting design. It is a serious build, and it is limited by the wrong physics for Mach 1. Patel 2026
What is actually limiting it
At 590 km/h the advance ratio is about J = 2.2 and the blade sections are near zero lift, so thrust only comes back if rpm keeps climbing. The binding limits are high-advance-ratio propeller thrust, drag area, and motor thermal margin. A 900 KV motor swap was already rejected on energy and thermal grounds, and aggressive rewinding did not fix the system. Patel 2026
Hall-effect and ion thrusters are not an option. They produce mN to N of thrust, carry their own propellant, need high-voltage supplies, and are not air-breathing, so they cannot push a 25 kg aircraft through Mach 0.8 to 1 and would not qualify anyway.
| Option | Evidence-based upside | Main problem |
|---|---|---|
| Custom high-pitch prop | Largest single lever in the brief. Higher pitch moves thrust and efficiency toward higher advance ratio. One CFD-plus-optimization study raised cruise efficiency from 82.3% to 87.1%; multi-point design cut required power about 4% at high speed versus a commercial APC prop. Patel 2026Czyz 2021Wu 2024Sanchez-Carmona 2025 | New blade design, spin rig, validation at very high rpm. |
| EDF redesign | Compact, high thrust per diameter, possible high-speed integration advantages. Gains over open rotors are most consistent at low speed. Urbán 2023Espinoza Quesada 2025Niță 2026Li 2023Jedamski 2023Darrah 2020 | Effectively a new aircraft: new ducts and inlets, a new drag problem. |
| Hall / ion / plasma | None in the aircraft literature. | Not air-breathing; thrust far too low. |
The real answer from Query 3: stop iterating the quad. Fixed-wing removes the propeller advance-ratio ceiling and cuts drag, and nearly every small supersonic program is built around a micro gas turbine, not distributed electric propulsors. Patel 2026Jacob 2021Compton 2021Mizobata 2005 The Dart-80 in the airframe viewer above is an open-source EDF delta, shown as a reference for what an EDF fixed-wing layout looks like.
Custom turbofan and
generative design?
It can help, if the fan stays low-bypass and topology optimization goes into parts, not the whole jet.
Engine cycle
Afterburning gives more augmentation in low-bypass turbofans than in turbojets, because the mixed stream is cooler and keeps more free oxygen, but fuel burn climbs sharply. Fetahi 2021 Overall pressure ratio and turbine inlet temperature dominate afterburning cycle performance, so set those first. Williams 2023Ighodaro 2022 Use reheat briefly to cross the drag rise; it can get through Mach 1 faster despite the fuel flow. Fetahi 2021Ighodaro 2022Apke 2012 Keep bypass very low; in the wet cases studied, thrust-specific performance falls as bypass ratio rises. Fetahi 2021Goulos 2018
Scaling is the trap. Very small turbojets already run only 5 to 10% overall efficiency, and one study argues a true tiny turbofan gets too complex and expensive to downscale cleanly. Schmidt 2021 Converting an existing micro turbojet with a fan, possibly geared or CVT-driven, is still a major propulsion project. Large 2018Kelly 2020 Curved after-bodies did not beat simple conical ones by much, though they can shorten length. Goulos 2018
Where topology optimization pays
| Part | Method | Reported result |
|---|---|---|
| Rotor blisk / disk | CFD-FEM topology optimization for mass and forced response Bandini 2024 | ~32% lighter |
| Turbine blade interior | Graded lattice / TPMS core Alkebsi 2021 | 33-40% lighter 26-49% less stress |
| Engine mounts, fittings | Regulation-constrained TO Munk 2019 | Min-weight parts |
| Whole wing | Gigavoxel morphogenesis Aage 2017 | only 2-5% |
Across these papers, topology optimization reliably cuts weight and can move resonances away from forcing frequencies, and additive manufacturing makes the shapes buildable. Munk 2019Bandini 2024Berrocal 2018 Point it at the rotor, hot-section internals, mounts and weight-critical structure. Fetahi 2021Shea 2024Boccini 2019 The bigger Mach 1.1 gains still come from cycle matching, nozzle control, diffuser losses and local mass cuts. Minato 2005Goulos 2018
Buy the hot core.
Make everything around it.
CNC, metal printing and PA6-CF, PC-CF and PEEK printing are real advantages, but the literature says to aim them at the integration, not the rotating hot section.
Buy
- The core turbojet, with ECU support and research precedent Prisăcariu 2026Catană 2024
- Servos, avionics, standard hardware (excluded from the amateur-built majority by the rules)
Hot sections need Inconel or refractory steels, and durability problems sit in thermal fatigue, combustion stability and manufacturing repeatability. Prisăcariu 2026 Metal AM fatigue is hard to qualify because defects, residual stress and anisotropy drive life. Sanaei 2020Javidrad 2024Liu 2023
Make
- Inlet lip and intake duct
- Engine mounts and nozzle shrouds
- Fuel tank internals, avionics trays, fairings
- Experimental add-ons: ejectors, afterburners, diffusers, manifolds, never compressor or turbine rotors Silivestru 2025Wattenbarger 2025
| Printed polymer | What the papers found | Use it for |
|---|---|---|
| PA6-CF | About 96 MPa tensile in favorable XY builds, but interlayer strength stays in the high-20 MPa range. Dul 2021 | Cool structure, oriented with the load |
| PC-CF | Better thermal stability and conductivity, still process-sensitive and orthotropic. Gupta 2022Gupta 2020 | Cool structure |
| PEEK | Strong fatigue for a printed polymer, but higher temperature and frequency cut life through defects and self-heating. Alhashmi 2025 | Warm-adjacent brackets, not hot hardware |
| All FFF composites | Anisotropic; properties depend on print direction and process; temperature or cyclic self-heating degrades fatigue. Goh 2020Peng 2020Gupta 2020 | Design for the weak axis |
Proving it, and
surviving it twice.
Winning needs a Mach number the judges accept and a vehicle that comes back through the transonic range intact, twice in one day.
Right inside the window you must cross. Safronov 2023 Active flutter suppression has extended margin past predicted onset in flight tests, but on far slower demonstrators. Vanek 2025 Clear margins with ground vibration testing, engine-on vibration tests and stepwise envelope expansion. Heidersbach 2023González-Cabrera 2026Sampath 2025
Design rule: spend as little time as possible in the drag-rise and flutter window. A strong core plus a short burst fits that better than a delicate mode switch in the same range. Liu 2024Xi 2023
Transonic and supersonic airspeed calibration depends on precise pressure and temperature data. Uncertainty drops with careful onboard calibration, reverse-heading logic and temperature-aware pressure correction. Zolovcik 1950Cooper 2014 Calibrate pitot-static with GPS-based methods before any high-speed claim, because installed pressure errors can corrupt Mach. Foster 2010Jurado 2024Zhang 2024
Engine bench too: measure thrust, fuel flow, EGT, pressure ratio and transient response, since catalog thrust is not enough. Rhee 2020bFernandes 2024Chachurski 2024Ezgin 2025
Nine steps
to a Mach 1.1 attempt.
The staged path the successful programs follow: bench first, slow prototype second, supersonic last. Apke 2012Shea 2024Lee 2025García 2024Mizobata 2014
Buy engine
P300 class minimum
Model thrust margin
Installed, not catalog
Shape airframe
Area rule, inlet with fuselage
Bench test
Engine, afterburner, diffuser, nozzle
Low-speed prototype
Check handling vs models
Full-scale taxi
Self-powered takeoff
Subsonic flight
Air-data calibration
Transonic build-up
Flutter margins, stepwise
Mach 1.1 attempt
Two flights, one day
The build the evidence points to
Single engine, fixed wing, thin swept wing, slender area-ruled fuselage, and add-on ejector or afterburner only after the dry engine is characterized. Dalman 2021Rhee 2020bMizobata 2016 Most published small-supersonic programs are still at design, wind-tunnel or low-speed prototype stage, so this is a plan with a clear risk, not a recipe.
Can a commercial 300 to 400 N engine keep enough net installed thrust, after inlet, nozzle and drag penalties, to hold a compliant self-launched Mach 1.1 segment at the final takeoff mass?
Sources and
3D credits.
Every paper the Consensus thread cites, with its Consensus link. Patel 2026 is the uploaded SQ-2 technical brief. One paper (Wang) appears in the thread only as "+1 more" with no year.
3D models
All meshes are the original authors' geometry. I converted STEP and STL to GLB, decimated with meshoptimizer, and quantized positions. On the KJ66 assembly, six small stray bodies that sit far off the engine axis were removed. On the X-59, the flat triangular print-support web under the belly (a 3D-printing aid, not part of the aircraft) was removed. Nothing is procedural or invented. Materials and colors were assigned for this page.
| Model | Author | Source | License | File here |
|---|---|---|---|---|
| KJ66 micro turbojet, full assembly (exploded layout) | Akash Jangid | github.com/Akash-9070/Turbo3D | MIT | models/kj66-assembly.glb 740 KB |
| KJ66 compressor wheel (impeller) | Akash Jangid | github.com/Akash-9070/Turbo3D | MIT | models/kj66-compressor-wheel.glb 55 KB |
| KJ66 combustion chamber casing | Akash Jangid | github.com/Akash-9070/Turbo3D | MIT | models/kj66-combustion-chamber.glb 265 KB |
| NASA X-59 (printable model) | NASA ARMD; Zachary Gwennap and Liam Brinton (NASA interns) | www.nasa.gov/stem-content/x-59-3d-printing/ | NASA media, public domain in the U.S. | models/nasa-x59.glb 514 KB |
| Dart-80 80 mm EDF delta jet | Caio Cerda (caioplcerda) | github.com/caioplcerda/dart-80 | CERN-OHL-S-2.0 | models/dart80-edf-delta.glb 452 KB |
| SQ-2 v3 speed quadcopter, 74-part CadQuery assembly (exploded) | Aayushya Patel | speed-quad repo (not public) | Own design | models/sq2-v3.glb 1402 KB |