Resonant Momentum Architecture (RMA)

Resonant Momentum Architecture: Technical Concept Paper

Resonant Momentum Architecture (RMA)

A cavity-stored, beam-assisted propulsion architecture for multi-modal spacecraft · Technical concept paper · Draft v1.0 · Status: conceptual, unvalidated

Abstract

The Resonant Momentum Architecture (RMA) is a propulsion concept built on one principle: electromagnetic energy carries momentum, and a well-designed system can collect that energy, hold it in a controlled state, and release it through a shaped port so that as much of that momentum as possible ends up in the spacecraft. Resonance governs where and how the energy is held, coupling governs when it leaves, and aperture geometry governs where it goes.

The architecture has three configurations. Configuration A is an onboard photon engine that releases stored electromagnetic energy as a directed pulse. Configuration B feeds that energy into magnetically confined plasma, so the exhaust is fast particles as well as light. Configuration C moves the energy store off the craft: a relay station transmits a pulse to a shaped momentum receiver. Each craft can carry several of these alongside chemical and electric propulsion, and the environment selects the active mode.

This paper sets out the physical principles, the design of each configuration with worked numbers, energy collection and thermal management, multi-modal operation, risks and a staged development path. It is checked against existing flight and laboratory evidence (Appendices A and B). The main findings are that the underlying physics is sound, that Configurations B and C have the strongest supporting evidence, and that an onboard photon-only engine faces severe limits in stored-energy lifetime, field strength and thrust per watt.

1. Introduction and design philosophy

1.1 The problem

Spacecraft meet very different resources as they travel. Near a star sunlight is abundant. Near a planet there is an upper atmosphere and a magnetic field. In the outer solar system energy is scarce and the environment offers almost nothing to push against. No single propulsion technology suits all of these. Chemical rockets give high thrust but consume propellant quickly. Electric thrusters are efficient but slow, and need both power and propellant. Solar sails need no propellant but produce very small forces and depend on light being available.

The RMA treats propulsion as a problem of managing energy and momentum as resources: collect them where they are plentiful, store them in whatever form suits the job, and convert them to thrust in the way the local environment allows.

1.2 Central principle

Energy amplification: no. Momentum-transfer efficiency: yes. The RMA does not create energy and does not beat p = E/c. Its aim is to get very good at delivering the momentum already carried by electromagnetic energy to the spacecraft, in a controlled direction. Resonance controls accumulation and coupling, receiver geometry controls capture and directional transfer, and an external mirror makes genuine photon recycling possible.

1.3 Origin of the concept

The concept began with a resonance analogy. When a plate is driven at a chosen frequency, rice on its surface collects in a stable pattern, because each frequency excites a particular vibration mode with its own arrangement of nodes and antinodes. An electromagnetic cavity behaves the same way: each resonant frequency excites a specific field pattern, and the pattern determines where the energy sits and which openings it can reach. The RMA turns that observation into an engineering question: can the pattern be chosen so that stored field energy leaves through one port, in one direction, on command?

1.4 The three configurations

ConfigurationEnergy storeExhaustPrincipal advantage
A: onboard photon engineOn the craftDirected lightNo propellant, works anywhere
B: plasma hybridOn the craftHeated plasma (and residual light)Orders of magnitude more thrust per joule
C: relay-beamed receiverAt a stationNone on the craft; momentum comes from the beamLight, simple vehicle; very high delivered energy

2. Physical principles

2.1 Momentum of radiation

A photon of energy E carries momentum p = E/c. A surface that absorbs the light receives this momentum once. A surface that reflects it at normal incidence receives it twice, because the photon reverses direction. For light of intensity I on area A, the force is I·A/c if absorbed and 2·I·A/c if perfectly reflected. At 1 AU, sunlight has an intensity of about 1,361 W/m² [1], which gives a radiation pressure of roughly 4.5 µPa if absorbed and 9 µPa if reflected. At oblique incidence the useful force falls with the cosine of the angle of incidence squared for a flat mirror, which is why sail and receiver geometry matter.

Absorbed: F = P / c · Reflected (normal): F = 2P / c
Impulse from energy E: Δp = E / c (absorbed), 2E / c (reflected)
Velocity change: Δv = E / (m·c)

2.2 Thrust per watt and exhaust velocity

For a conventional exhaust of mass flow ṁ and velocity v, the thrust is F = ṁ·v and the jet power is ½·ṁ·v², so the ideal thrust per unit power is 2/v. A slow exhaust gives a lot of thrust per watt but consumes propellant quickly, and a fast exhaust does the opposite. This is the trade that separates chemical, ion and plasma engines. Light is the limiting case: its exhaust velocity is c, and the thrust per unit power is 1/c, about 3.3 nN/W.

ExhaustSpeedIdeal thrust per watt
Plasma or ablated propellant3 km/s≈ 670 µN/W
Ion or plasma thruster30 to 50 km/s≈ 40 to 70 µN/W (VX-200 measured 29 µN/W [11])
Lightc3.3 nN/W (6.7 nN/W if reflected once)

The consequence is that light is the worst exhaust per watt and the best per kilogram of propellant, since it needs none. Whether that trade pays depends on whether propellant or energy is the scarcer resource for a given mission.

2.3 Resonant cavities: modes, quality factor and stored energy

A cavity supports standing electromagnetic waves only at discrete frequencies, called modes. Each mode has a fixed spatial field pattern set by the cavity geometry. For a rectangular metal cavity of sides a, b and d the resonant frequencies follow f = (c/2)·√((l/a)² + (m/b)² + (n/d)²) for integer mode numbers, and each combination has its own arrangement of field maxima and nodes. This is the electromagnetic counterpart of the Chladni pattern on the plate.

Two quantities describe how well a mode holds energy. The quality factor Q is the ratio of stored energy to the energy lost per radian of oscillation, Q = ωU / P_loss, so the stored energy decays with time constant τ = Q/ω. The coupling coefficient β = Q₀/Q_ext compares the wall losses (Q₀) with the losses through the output port (Q_ext). The loaded quality factor obeys 1/Q_L = 1/Q₀ + 1/Q_ext, and the fraction of the stored energy that leaves through the port is β/(1+β). A port that is much lossier than the walls (β ≫ 1) is called over-coupled, and this is the regime a thruster needs: nearly all the energy leaves as useful output and very little becomes wall heat.

Superconducting niobium cavities already reach Q₀ of about 3 × 10¹⁰ at gigahertz frequencies [16], so the wall term can be made very small. The catch, developed in Section 4.3, is that a high Q gives a long decay time only when the frequency is low, and that fields, not Q, are what limit how much energy a cavity can hold.

2.4 Output coupling and cavity dumping

Lasers use a technique called cavity dumping to release stored energy in a short pulse. The cavity is kept highly reflective while energy builds up, then a switch rapidly raises the coupling so that the stored field leaves in roughly one round-trip time, 2L/c for a cavity of length L. The RMA uses the same idea: the port is effectively closed during accumulation and opened on command. The switch itself is a design problem in its own right, and at the power levels considered here it becomes the component with the least precedent.

2.5 Radiation and plasma

Plasma responds strongly to electromagnetic fields. Its natural oscillation frequency, the plasma frequency, is ω_p = √(n·e²/(ε₀·mₑ)) for electron density n. Waves below ω_p are reflected and waves above it propagate, so a dense plasma can act as a mirror, and a wave at the critical density is absorbed strongly. For 10 GHz radiation the critical density is about 10¹⁸ m⁻³, which is the density reported near the exit of the VX-200 thruster [11].

Energy moves from field to particles by several routes. In collisional absorption (inverse bremsstrahlung), electrons driven by the field collide with ions and convert field energy to heat. In resonant heating, a wave tuned to a natural frequency of the plasma, such as the ion cyclotron frequency ω = qB/m, transfers energy efficiently to the ions. In the ponderomotive effect, intense fields push charged particles away from regions of high intensity. Once the plasma is hot, a magnetic nozzle converts the random motion of the particles into directed flow along the field, using the conservation of the magnetic moment μ = m·v⊥²/(2B) as the field weakens, and the plasma detaches from the field lines downstream.

2.6 Conservation laws and the system boundary

Momentum is conserved for any closed system. Light bouncing between two mirrors on the same craft pushes each mirror in turn, and the forces cancel: internal reflection redistributes momentum inside the system and adds none to it. Net thrust appears only when momentum crosses the system boundary, either by radiation leaving the craft or by radiation arriving from outside. This is why the onboard engines derive all their thrust from what the port emits, and why a station mirror facing the craft changes the picture: the mirror lies outside the craft’s boundary, so the same photons can hand momentum to the craft on each pass. It is also why a sealed cavity cannot produce net thrust, which is what the EmDrive tests found [10].

3. System architecture

3.1 Functional chain

CollectSunlight, nuclear or beamed power; optional plasma scoop
StoreBulk electrical or magnetic storage
ConvertPower electronics and RF or optical source
ConditionResonant cavity: mode, timing, field build-up
ReleasePort, optics or magnetic nozzle

Every stage has an efficiency below one, and the losses become waste heat that the thermal system must reject (Section 7.3). The overall efficiency is the product of the stage efficiencies, so weak stages dominate.

3.2 The three controls

ControlWhat it setsPhysical mechanism
Resonant frequencyWhich mode holds the energy, and whereCavity eigenmodes
Stored field energyHow much momentum is availableU, limited by breakdown and quench fields
Coupling apertureWhen the energy leaves and where it goesβ (timing), aperture size and shape (direction)

This separation is the design’s main organising idea. Frequency decides where and how the energy sits, coupling decides when it leaves, and aperture geometry decides where it goes. Because the three are largely independent, each can be tuned without redesigning the others, and switching between modes offers thrust vectoring with no moving parts.

3.3 Pulsed operation and power arithmetic

The system works in pulses. If E_pulse is released in a time t_pulse, the peak power is E_pulse/t_pulse, while the average power is E_pulse × f_rep. The energy store must supply E_pulse per cycle, and it recharges at the rate the collector allows, so the recharge time is E_pulse divided by the collected power. The benefit of pulsing is that peak power can far exceed the average power collected. A second benefit applies on elliptical orbits: a short, strong burn near periapsis extracts more benefit per joule from the Oberth effect than the same energy spread over the whole orbit.

4. Configuration A: onboard photon engine

4.1 Concept

Electrical energy is converted to radio-frequency or optical power, fed into a resonant cavity, and released as a directed pulse through an output port. No propellant is used. The exhaust is light, so the design inherits the 2.2 trade: no propellant, but the lowest thrust per watt of any exhaust.

4.2 Cavity design

The cavity has to combine three properties that pull against each other: a high Q so that little energy is lost to the walls, a mode whose field pattern couples cleanly to one port, and a structure that survives large field energy densities. The choice of frequency matters. Lower frequencies give larger cavities and lower stored-energy density but a shorter decay time for a given Q; higher frequencies allow compact optics and finer beam control. Mode purity is important because a multimode field leaks diffusely through any opening, while a single mode leaves as a coherent beam.

4.3 The stored-energy lifetime constraint

The decay time τ = Q/ω limits how long a cavity can hold energy. At 10 GHz (ω ≈ 6.3 × 10¹⁰ rad/s) a cavity with Q₀ = 3 × 10¹⁰ has τ ≈ 0.5 s. At an optical frequency of 6 × 10¹⁴ Hz and Q = 10¹⁰, τ is only a few microseconds. Holding an energy U continuously dissipates U/τ, so 300 GJ in the 10 GHz case would dissipate about 600 GW as heat, and the field required would be far beyond the low-field conditions at which those Q values are measured.

The consequence is that the cavity cannot be the long-term store. It is a fast buffer that conditions energy over seconds or less. The bulk energy must sit in batteries, capacitors or superconducting magnetic storage, and it has to be delivered to the cavity fast enough to fill it within about one decay time. Filling 300 GJ in 0.1 s needs a feed of about 3 TW. The earlier idea of collecting for an hour and releasing in a second therefore applies to the bulk store, not to the cavity, and the cavity does not reduce the peak power the store must deliver.

4.4 Output port and release dynamics

The port sets the release time and the emission direction. Beam divergence for an aperture of width D is roughly θ ≈ λ/D, and the axial momentum lost is about θ²/2. A 3 cm wave through a 1 m aperture diverges by about 30 mrad and loses roughly 0.05% of axial momentum, while a plain hole in a multimode cavity radiates diffusely and delivers about two thirds of the ideal thrust. Release time is set by the loaded quality factor, roughly Q_L/ω, and cannot be shorter than about one light round trip.

Release timePeak power (300 GJ)Peak forceNote
1 s300 GW≈ 1,000 NGentle on the structure
1 ms300 TW≈ 1 MNNeeds strong mounts and mirrors
≈ 140 ns≈ 2 × 10¹⁸ W≈ 7 GNOne light round trip across a 21 m cube; up to ≈ 700,000 g on a rigid 1,000 kg craft

Total impulse is the same in every row, about 1,000 N·s, so the velocity change is identical. Only the peak load differs, which makes coupling strength a design variable to tune against what the structure can survive.

4.5 Sizing and field limits

Craft mass and ΔvEnergy to emit (E = m·c·Δv)Equivalent
1,000 kg, 1 m/s3 × 10¹¹ J≈ 83 MWh
1,000 kg, 100 m/s3 × 10¹³ J≈ 8.3 GWh
10 kg, 1 m/s3 × 10⁹ J≈ 830 kWh

The radiation pressure inside the chamber depends on the volume that holds the energy. For 300 GJ the pressure is about 100 GPa in 1 m³ and about 10 MPa in 10,000 m³, comparable to a rocket combustion chamber. The corresponding peak electric field in the larger chamber is about 2.6 GV/m. Today’s superconducting cavities operate at accelerating gradients of tens of MV/m [16], so the design needs fields more than an order of magnitude beyond current practice. For comparison, the most energetic pulsed laser system built, the National Ignition Facility, delivered 2.2 MJ in a single shot [15], about 10⁵ times less than 300 GJ.

4.6 Assessment

Configuration A is physically valid but energy-hungry. A 1,000 kg craft needs hundreds of gigajoules for a single metre per second. Near a star a solar sail of the same size reaches that velocity in a few hours without any of this machinery (Section 10), and the field strengths and dump power sit well beyond demonstrated practice. The configuration makes most sense where energy is limited but propellant is not allowed, and at small scale: formation flying and precision pointing, where the required impulse is tiny and photon thrust offers very fine control, as photonic laser thruster work aims for [14]. It also serves as the reference case against which B and C are judged.

5. Configuration B: plasma hybrid

5.1 Concept

Instead of releasing the cavity’s energy as light, the hybrid delivers it to a small amount of propellant, which becomes hot plasma and leaves through a magnetic nozzle. The exhaust is now made of particles, so the thrust per joule follows the 2/v law of Section 2.2 and improves by orders of magnitude over light.

Exhaust (1,000 kg craft, 1 m/s)Energy neededPropellant
Light3 × 10¹¹ JNone
Plasma at 3 km/s≈ 1.5 × 10⁶ J≈ 0.33 kg

The two options differ by a factor of about 200,000 in energy, at a propellant cost of a third of a kilogram. This is what removes the storage and field-strength problems of Configuration A.

5.2 Energy path

Bulk storage feeds an RF or microwave generator, which drives a coupler that delivers power to the plasma. Electrical energy is converted to field energy, then to plasma heat, then to directed kinetic energy in the nozzle. The overall efficiency is the product of the generator, coupling, heating and nozzle efficiencies. The VX-200 prototype measured 72% overall efficiency at 200 kW in a vacuum chamber, with a specific impulse near 4,900 s and thrust of 5.8 N [11].

5.3 Plasma source, heating and confinement

In the VASIMR design, a helicon source ionises the gas and a second stage heats the ions with radio waves at their cyclotron frequency inside a superconducting magnet, with no electrodes exposed to the plasma [11]. Magnetic confinement holds the hot plasma off the walls, which protects the structure and avoids electrode erosion. The trade-off is the mass and power of the magnets and the difficulty of containing plasma instabilities.

5.4 Nozzle, propellant and pulsing

The magnetic nozzle converts the plasma’s thermal energy to axial flow, and the plasma must detach from the field lines to produce net thrust. Propellant choice affects ionisation cost, storage and exhaust speed. Argon is used in the VX-200 tests [11]. In low orbit, the atmosphere itself can supply propellant, as the ESA air-breathing thruster tests show [13].

The hybrid can also run in pulses. A pulsed plasma thruster stores energy in a capacitor and discharges it across a propellant surface, and has flown since 1964 [12]. Flight units use pulse energies of roughly 5 to 55 J, so an impulse of 1,000 N·s at 3 km/s corresponds to about 1,000 pulses of 1.5 kJ or many more smaller ones. A resonant feed adds mode selection and controlled timing to this pulsed model.

5.5 Assessment

Configuration B is the design with the strongest supporting evidence: capacitor-discharge thrusters are flight-proven and RF plasma heating with magnetic nozzles has been tested at 200 kW. Its costs are propellant, magnet mass and the thermal load of the conversion chain. Its advantage over an ordinary ion or plasma thruster has to come from the resonant heating stage, and that claim needs testing.

6. Configuration C: relay-beamed momentum receiver

6.1 Concept

A station stores energy over a long time and transmits it as a pulse or a continuous beam. The craft carries a shaped receiver that captures the beam and converts its momentum to thrust. The energy store, collector and chamber remain at the station, where mass matters far less. Beamed-energy sails have been studied for decades and are the basis of Breakthrough Starshot [5]. Small-scale beaming in space has been demonstrated, with a flexible phased-array transmitter steering microwave power without moving parts [7].

6.2 Station and beam

The station comprises a power source, bulk storage, a generator and a transmitting aperture. Beam direction and focus come from the aperture: a phased array shapes the wavefront electronically, and the mode and aperture design of Section 4.4 carries over. Diffraction limits how far a beam stays narrow. The spot size at range L is about λL/D_t, so the beam matches a receiver of width D_r out to L ≈ D_t·D_r/λ.

Wavelength1 km transmitter, 100 m receiver
3 cm microwave≈ 3,300 km, so short local hops only
1 µm optical≈ 10¹¹ m (about 0.7 AU)

Relay lenses or mirrors along a route can refocus the beam and extend the range.

6.3 Receiver optics and thermal limits

A flat mirror at normal incidence gives 2E/c, the ceiling for a single reflection, so 1 m/s for a 1,000 kg craft needs 150 GJ of intercepted energy. A continuous 1 GW beam would deliver that in 150 s at a thrust of about 6.7 N. The receiver’s job is to approach that ceiling: capture the beam, turn oblique reflections into reversed ones, and stay on target. Concave or conical beam-riding shapes give restoring forces that keep the receiver centred in the beam, and segmented or funnel geometries increase the effective capture area. Diffraction is not fixable at the receiver, since light that misses is lost.

Heat is the second limit. With reflectivity R, the absorbed power is (1 − R) of the beam. For a 1 GW beam and R = 0.9999, about 100 kW is absorbed, which over 10,000 m² is a manageable 10 W/m². Reflectivity therefore sets how much power the receiver can accept, and high-reflectivity multilayer dielectric coatings are central to the design.

6.4 Photon recycling

A funnel geometry can approach the 2E/c limit but cannot exceed it, and internal bounces inside the craft cannot either, because the net transfer equals incoming momentum minus outgoing. To go beyond it the returned light must hit the craft again, which requires an external mirror at the station. Sail and station mirror then form a resonant cavity, and a static cavity of mirror reflectivity R gives roughly 1/(1−R) passes. Photonic laser thruster experiments report recycling between two mirrors, with amplification factors of thousands [14], though the results come mainly from one group.

The limit is dynamics. As the craft recedes, the round-trip time grows and the mirror spacing must be held to a fraction of a wavelength for the whole storage time. An active cavity with a gain medium inside tolerates mirror motion better than a passive one, which is the approach of the photonic laser thruster. Recycling therefore suits the acceleration phase near the station and not the coast.

6.5 Assessment

Configuration C delivers the highest energy to a craft with the simplest vehicle. Its costs are infrastructure, beam pointing over long ranges, and receiver heating. It suits a repeatable route, such as a transfer corridor, better than a one-off mission.

7. Energy collection, storage and thermal management

7.1 Collection

Sunlight is the main source near a star: 1,361 W/m² at 1 AU [1], or about 400 W/m² electrical at 30% efficiency, falling with the square of distance to about 50 W/m² at 5 AU. Farther out, a compact nuclear source is needed, such as radioisotope generators or reactors. Propellant can be collected as well: the solar wind gives about 3 × 10⁻¹⁵ kg/m²/s at 1 AU [2], roughly 0.1 kg per year per km² of scoop, which is enough for a small hybrid reserve over years but not for fast cycling. The upper atmosphere in low orbit is far richer, at the cost of drag [13]. The vacuum itself provides no net energy (Section 9).

A 10,000 m² array yields about 4 MW at 1 AU, which recharges the 300 GJ photon case in about 21 hours and the 1.5 MJ hybrid case in under a second.

7.2 Storage

The bulk store has to hold the energy between collection and use and deliver it at the peak rate the cavity or plasma stage needs. The main candidates trade energy density against power density. Batteries store the most energy per kilogram but deliver power slowly; capacitors and superconducting magnetic storage deliver power very fast but hold less energy per kilogram. As a scale reference, 83 MWh, the energy for the 1 m/s photon case, would need on the order of 400 tonnes of batteries at typical lithium-ion energy densities. Hybrid stores, with batteries for bulk and capacitors for peak delivery, are the likely answer.

7.3 Thermal management

Every conversion stage produces waste heat, and in vacuum the only way to reject it is by radiation, at a rate εσT⁴ per unit area. At 400 K with emissivity 0.9, a radiator sheds about 1.3 kW/m² on one side, so rejecting 1 MW needs roughly 770 m². Cavity and coating losses add localised heating on top of this. Thermal design is a first-order constraint on all three configurations, and often sets the practical power limit before any physical law does.

8. Multi-modal integration and operations

8.1 The propulsion ecosystem

SubsystemRole
Photon or plasma engine (A and B)Autonomous, propellant-light or propellant-free thrust on locally stored energy
Momentum receiver (C)Accepts beamed momentum on network routes, with optional recycling near a station
Chemical propulsionShort, high-thrust events: emergency acceleration, landing, avoidance. Solid fuel with oxidiser, or a hybrid of solid fuel with stored oxidiser
Electric propulsionIon or plasma thrusters on a small propellant mass, sharing the energy store
Attitude controlReaction wheels, cold gas or electromagnetic torquers, so orientation does not need the main engines
Energy collection and storageSolar collectors or other receivers, with batteries, capacitors or superconducting magnetic storage
Thermal managementRadiators and heat transport for every conversion stage
Navigation and controlSchedules collect, store, resonate, couple and release, and selects the active mode

8.2 Mode selection by environment

SituationPrimary modeEnergy source
Inside the networkBeamed receiver (C), with local propulsion for manoeuvringRelay stations
Near a star, off the networkSolar sail or hybrid (B)Sunlight
Low orbitAir-breathing electric or hybridSunlight and atmosphere
Far from a starHybrid (B) on nuclear power, chemical for emergenciesOnboard source
High-thrust eventChemicalStored propellant

8.3 Operating inside and outside the network

All craft can operate both inside and outside the network. Infrastructure extends capability and does not determine whether a craft can fly. Inside, the craft receives beamed energy, uses its receiver and recycling, takes gravity-assist routes where useful, and recharges its stores. Outside, it collects energy itself, stores it, and uses its onboard engines. The network grows with exploration: a craft can seed a new node by carrying or later building the store and transmitter, which are the heavy parts of a station, so a node does not appear instantly.

The cost is mass and complexity. Every added system adds both, so shared subsystems (energy store, thermal, control) are essential, and mission analysis decides what each craft carries.

9. Vacuum and quantum-field considerations

The word vacuum has layers. Ordinary vacuum has very few atoms. Interplanetary and interstellar space is sparse but holds sunlight, the solar wind, magnetic fields and dust. The quantum vacuum is the lowest energy state of the fields, and a perfect vacuum is only a mathematical limit. The interplanetary layer is a real resource that sails and scoops use. The quantum layer is the medium through which light propagates, and it has no energy or momentum to give, so it cannot serve as propellant.

The chamber does not confine particles in the ordinary sense. It confines an excitation of a field: a standing mode holding an enormous number of photons, about 10³⁰ at optical frequencies for 300 GJ, which behaves as a classical field. The walls, mirrors and coupler shape that mode and decide where it leaves.

Several vacuum effects are real but not useful for thrust. The Casimir force is static and cannot power a cycle. The dynamical Casimir effect converts vacuum fluctuations into real photons using a moving boundary, and has been observed in a superconducting circuit [8], but the energy comes from the boundary motion and is tiny. Schwinger pair creation needs about 1.3 × 10¹⁸ V/m, which has not been reached in the laboratory [9]. Claimed propellantless cavity thrusters have not survived careful testing [10]. Other particle exhausts fare no better: neutrinos interact too weakly and quarks are confined, so photons and heavy charged particles remain the practical choices.

10. Performance comparison and mission fit

SystemThrust per wattPropellantEvidence
A: photon engine3.3 nN/W (ideal, one pass)NoneNo precedent at scale
B: plasma hybrid≈ 30 to 500 µN/W (3 to 50 km/s exhaust, ~70% efficient)Grams to kilogramsGround-tested, pulsed units flown
C: beamed receiver6.7 nN per watt of intercepted beam (more with recycling)NoneSails flown, beaming at small scale
Solar sail near a starFree sunlight: ≈ 0.09 N on 10,000 m² at 1 AUNoneFlown [3, 4]
Ion thruster≈ 40 µN/W (NEXT-class) [11]Xenon or similarFlown

The comparison shows where each option wins. A sail beats photon engines near a star: 10,000 m² of reflective sail reaches 1 m/s on 1,000 kg in about 3 hours. The hybrid wins where propellant is available and energy is limited. The beamed receiver wins where a network exists and a simple vehicle matters. The onboard photon engine wins only where propellant is prohibited and the impulse is small.

11. Risks and failure modes

RiskEffectMitigation
Cavity breakdown or quench at high fieldUncontrolled release; 300 GJ is about 70 tonnes of TNT equivalentKeep stored energy far below limits; use B or C at large scale
Off-axis emissionDirect loss of thrustMode purity, collimating optics, over-coupling
Extreme peak force on fast dumpsMirror and mount damageWeaker coupling to lengthen the release
Waste heatOverheating; limits average powerRadiator area, high-Q design
Plasma instabilities and wall erosion (B)Scattered light, lost efficiencyMagnetic confinement, tuned heating
Slow propellant supply from the solar windLong refill timeStored reserve; use atmosphere where available
Receiver heating and beam pointing (C)Damage or lost energyHigh-reflectivity coatings; relay optics, beam-riding shapes
Multi-modal mass and complexityPayload penaltyShared subsystems; mission-driven selection
A sealed cavity claiming thrustViolates momentum conservationEmission path must be open whenever thrust is claimed

12. Development roadmap

StageActivitySuccess criterion
0. AnalysisFix a target mission and select configuration B or C; model chamber, plasma and beamSizing and thermal budgets close on paper
1. Bench testsTorsion-pendulum tests in vacuum of a resonantly fed pulsed plasma discharge; cavity-dump tests; two-mirror recycling replicationMeasured momentum per joule matches prediction; independent replication of recycling
2. Ground integrationCombined feed, plasma and nozzle in a thermal-vacuum chamberEfficiency and thrust at target power
3. Small flight demonstrationCubesat-class pulsed hybrid with a resonant feedIn-orbit thrust and efficiency data
4. Relay demonstratorLow-orbit beam from a station to a reflector on a craftMeasured beamed momentum against the 2E/c prediction

At every stage the results are benchmarked against a solar sail and a conventional electric thruster, since the architecture has to beat them to justify its complexity.

Appendix A: Existing projects and results

Verified against the sources in the References list on 24 September 2026. Status can change, so check the primary source before relying on an entry.

ProjectResult or statusRelevance to the RMA
LightSail 2 (Planetary Society)2019: controlled solar sailing in Earth orbit, orbit raised by about 2 km in the first days using only light pressure [3]Sunlight momentum works in flight; baseline for the receiver and sail comparison
IKAROS (JAXA)2010: first solar sail in interplanetary space; 20 m sail, passed Venus in December 2010; sail also carried thin-film solar cells [4]Flown sail plus onboard power; hybrid sail and electric propulsion was the stated goal
Breakthrough StarshotConcept: ~100 GW ground laser array, few-metre sail, minutes of push, roughly 10¹³ J for a gram-scale probe at 0.2 c [5, 6]. Independent critique on cost and realism [6]. Recent commentary describes it as on holdConfiguration C at extreme scale; diffraction, sail heating and beam pointing are the limits
Caltech SSPD-1 / MAPLE2023: first wireless power transfer in space by a lightweight flexible microwave array; beam steered by phase, with no moving parts, and detected on Earth [7]Relay transmitter and beam steering; small power (LED-scale), not GW
Photonic Laser Thruster (Bae Institute)Reported photon-thrust amplification of ~3,000 by bouncing light between two mirrors, 35 µN (2006 to 2007) [14]; later reports up to 3.5 mN with a 500 W laser and ~1,540 recycles over a few metres [14]Closest experimental match to external-mirror recycling; results come mainly from the inventor’s group
VASIMR VX-200 (Ad Astra)Ground tests: 72% efficiency, Isp about 4,900 s, 5.8 N at 200 kW total power, using helicon plasma and ion-cyclotron RF heating [11]Configuration B: radiation heating plasma, magnetic nozzle exhaust. I did not verify any later in-space test
Pulsed plasma thrustersFlew on Zond 2 in 1964; used for pitch control on EO-1; TRL 9 for attitude control and station keeping [12]Capacitor-discharge model for the pulsed hybrid; thrust and pulse energy are small
Air-breathing electric propulsion (ESA/Sitael RAM-EP)2018: ground firing that ingests air simulating 200 km altitude at 7.8 km/s; AETHER tests reported 7 to 20 mN at 1 kW [13]Propellant collection from a planet’s upper atmosphere
National Ignition Facility (LLNL)192-beam laser; record 2.2 MJ delivered on 30 October 2023 [15]Scale of pulsed light energy today: about 10⁵ times below the 300 GJ photon case
Superconducting RF cavities (LCLS-II, Fermilab)Q₀ about 3 × 10¹⁰ at 2 K and about 16 MV/m [16]High-Q cavities are real; their fields are far below the ~2.6 GV/m of the 300 GJ, 10,000 m³ chamber
Dynamical Casimir effect (Chalmers)2011: photons created from vacuum fluctuations by a moving boundary in a superconducting circuit [8]A possible photon source; tiny energy, not a thrust source
Electric solar wind sail (FMI, ESTCube)Concept proposed 2004 [17]. ESTCube-1 flew but its tether deployment was unsuccessful, so no in-space sail thrust was measured [17]Solar-wind sails remain unproven in flight
EmDrive-type cavity thrustersTU Dresden found apparent thrust came from magnetic and thermal artefacts, not the cavity [10]Negative evidence: sealed-cavity thrust is not supported
Schwinger pair creationCritical field ≈ 1.3 × 10¹⁸ V/m; not reached in the laboratory [9]Vacuum pair creation is not a usable propellant source

Appendix B: Evidence status of each element

ElementEvidence levelMain gap
Momentum from reflected light (sails)FlownThrust is tiny; scale-up
Beamed power in spaceSmall-scale flownGW power; beamed momentum on a craft
Beam-driven sail at scaleConcept studies onlyHuge arrays, sail heating, pointing
External-mirror photon recyclingLab, mostly one groupIndependent replication; a moving craft; scale
Pulsed capacitor-discharge thrustFlown (TRL 9)Low thrust; joule-scale pulses
RF plasma heating and magnetic nozzleGround-testedIn-space operation; mass and thermal budgets
Air-breathing propellant collectionGround-demonstratedIn-orbit demonstration
Solar-wind sailsConcept; first test unsuccessfulAny in-space thrust measurement
High-Q superconducting cavitiesProven at low fieldField strength about 10 to 50 times higher and energy about 10⁵ times higher than today’s practice
Onboard cavity-dump photon engine (configuration A) at GJ scaleNo direct precedentStorage, containment, peak loads
Vacuum-derived thrust; sealed-cavity thrustNone; refuted where testedConflicts with momentum conservation

Overall: the physics in the spec (p = E/c, momentum recycling, plasma heating, pulsed discharge) is well supported. The parts with real flight or lab evidence are sails, beamed power, pulsed and RF plasma thrusters and high-Q cavities. The specific onboard gigajoule photon engine has no precedent, and near a star the evidence favours a plain sail over it. The plasma hybrid (configuration B) and the relay-beamed receiver (configuration C) have the strongest supporting evidence.

Appendix C: Calculation notes

QuantityMethod and inputsResult
Photon energy for ΔvE = m·c·Δv; m = 1,000 kg, Δv = 1 m/s3 × 10¹¹ J (150 GJ if reflected, 2E/c)
Plasma sizingImpulse 1,000 N·s at 3 km/s: m = 0.33 kg; E = ½mv²≈ 1.5 MJ
Thrust per wattPhoton: 1/c. VX-200: 5.8 N / 200 kW. NEXT: 0.327 N / 7.7 kW [11]3.3 nN/W vs about 29 and 42 µN/W (ratio ≈ 10⁴)
Radiation pressureP ≈ u/3 for isotropic field; u = 300 GJ / volume100 GPa at 1 m³; 10 MPa at 10,000 m³
Field strengthu = ε₀E₀²/2 at 3 × 10⁷ J/m³≈ 2.6 GV/m
Peak loadRelease in one light round trip, 2 × 21.5 m / c; F = P/c≈ 140 ns, 2 × 10¹⁸ W, 7 GN
Recharge time1,361 W/m² × 30% × 10,000 m² = 4 MW [1]300 GJ in ≈ 21 h
Sail comparisonF = 2·I·A/c with I = 1,361 W/m², A = 10,000 m²≈ 0.09 N; 1 m/s in ≈ 3 h
Solar wind mass fluxn = 5 cm⁻³, v = 400 km/s, proton mass; typical n 3 to 7 cm⁻³ [2]3.3 × 10⁻¹⁵ kg/m²/s; ≈ 0.1 kg/yr per km²
Diffraction rangeL ≈ D_transmitter × D_receiver / λ3,300 km at 3 cm; 10¹¹ m at 1 µm (1 km, 100 m)
Schwinger fieldE_c = mₑ²c³ / (eħ) [9]1.3 × 10¹⁸ V/m

These are order-of-magnitude estimates for an ideal, lossless case. They are for sizing and comparison, not design values.

References

  1. NASA Goddard, Solar Irradiance Science (total solar irradiance 1361 W/m², SORCE and TSIS-1). earth.gsfc.nasa.gov
  2. NASA Technical Reports Server, “The Solar Wind” (about 7 ions/cm³ at 1 AU, 300 to 700 km/s): ntrs.nasa.gov/citations/19990071284. Slow and fast wind densities: arXiv:1708.07169
  3. The Planetary Society, LightSail 2 sail deployment: planetary.org. Mission success report: Universe Today
  4. Tsuda, Y. et al., “Achievement of IKAROS: Japanese deep space solar sail demonstration mission,” Acta Astronautica 82(2), 183 to 188 (2013): record. JAXA project page: global.jaxa.jp
  5. Parkin, K. L. G., Breakthrough Starshot system model: arXiv:1805.01306. Overview: Wikipedia (secondary). Status commentary: Centauri Dreams, March 2026
  6. Milchberg, H., “Challenges abound for propelling interstellar probes,” Physics Today (2016): physicstoday.aip.org
  7. Caltech, “In a first, Caltech’s Space Solar Power Demonstrator wirelessly transmits power in space” (2023): caltech.edu
  8. Wilson, C. M. et al., “Observation of the dynamical Casimir effect in a superconducting circuit,” Nature 479, 376 to 379 (2011): doi:10.1038/nature10561
  9. Schwinger critical field: LUXE collaboration, arXiv:2207.13510; strong-field QED at ELI-NP, arXiv:2307.09315 (field not yet reached experimentally)
  10. Tajmar, M. and Fiedler, G., “Direct Thrust Measurements of an EM Drive and Evaluation of Possible Side-Effects” (AIAA); Tajmar et al., “The SpaceDrive Project: First Results on EMDrive and Mach-Effect Thrusters” (2018), reported in The Register. Summary including the 2021 analysis: Wikipedia (secondary)
  11. Squire, J. et al., “VASIMR VX-200 thruster throttling optimization from 30 to 200 kW” (APS GEC 2012): abstract. Programme history: eoPortal. NEXT comparison (0.327 N at 7.7 kW): Wikipedia (secondary)
  12. Pulsed plasma thrusters: Wikipedia (secondary, Zond 2 in 1964); NASA Glenn, EO-1 PPT thrust testing: NTRS
  13. ESA, “World-first firing of air-breathing electric thruster” (5 March 2018): esa.int. AETHER project: IQM
  14. Bae, Y. K., photonic laser thruster: Laser Focus World (2007); later results summarised in Wikipedia (secondary). Independent replication not verified
  15. Lawrence Livermore National Laboratory, NIF records: lasers.llnl.gov; 2.2 MJ shot: LLNL news
  16. Gonnella, D. et al., LCLS-II cavity performance: arXiv:1411.1659; Posen, S. et al., arXiv:1812.03950; SRF R&D directions: arXiv:2204.01178
  17. Janhunen, P., “Electric sail for spacecraft propulsion,” J. Propulsion and Power 20, 763 to 764 (2004), and related papers: electric-sailing.fi; ESTCube-1 outcome: Wikipedia (secondary)

Appendix D: Design rationale log

QuestionConclusion
Can the resonant pattern itself become thrust?No. Resonance controls how energy is accumulated and where it exits; thrust comes from momentum leaving.
Can subatomic excitations (quarks, neutrinos, bosons) be used?Photons yes. Other bosons possible in principle. Neutrinos impractical (extremely weak interaction). Quarks confine into hadrons. Production and directed emission are separate problems.
Can it ignite as a controlled explosion with nothing extra?Yes as a cavity-dumped photon pulse. The port supplies direction. The failure mode is uncontrolled release at full stored energy.
Can stored electrical energy become a car-engine-style explosion?Yes, with a small working mass: a pulsed plasma thruster (capacitor discharge ablating propellant) is a flown example.
Can an electromagnetic explosion occur in vacuum without propellant?Yes: 1 GJ emitted as a directed pulse gives about 3.3 N·s. The limit is thrust per joule, not feasibility.
Is a contained chamber with an exhaust port the same dynamic in space?Yes. Chamber contains pressure, port channels it. The working fluid can be light, plasma or both.
Can radiation be combined with plasma to boost output?Yes. Resonant and collisional heating move energy from radiation into hot particles, raising thrust per joule.
Can everything be collected in real time in space?Energy: yes near a star. Propellant: a little from solar wind, more from upper atmosphere. Vacuum energy: no.
Can chamber shape and resonance push the energy out one aperture?Yes, through two levers: mode selection sets beam quality, aperture size sets divergence. Port coupling must dominate wall loss (Section 9).
Does a faster release change the thrust?It raises peak power and force in proportion, but total impulse is unchanged. The fastest dump is about one light round trip of the cavity.
Can the vacuum itself be propellant?Not directly. It has no energy or momentum to hand over. Sunlight, solar wind and magnetic fields in space can be used (Sections 7, 12).
Does “vacuum” mean nothing?No. Four layers, and the chamber confines a field excitation, not particles.
What are the three controls?Frequency sets where energy sits, coupling sets when it leaves, aperture geometry sets where it goes.
Can the energy store leave the craft?Yes: relay-beamed configuration C, with 2E/c per reflection but new infrastructure and receiver heating.
Can sail shape beat dispersion?Not diffraction. Shape aids stability and reflection angle; relay lenses extend range.
Can a funnel receiver recycle photons?It approaches the 2E/c ceiling. Going beyond needs an external station mirror forming a cavity.
Can an onboard mini-engine use internal reflection?Yes, for accumulation, shaping and timing. Net momentum comes from the photons that leave; internal reflection alone adds none.
What happens beyond the infrastructure network?The craft switches to the onboard tier on locally harvested energy, with conventional propellant as backup (Section 11).
Can one craft carry several propulsion systems?Yes: a multi-modal ecosystem (photon, chemical, plasma, attitude control) with shared energy, thermal and control subsystems, and the environment selecting the active mode.
Can craft seed new network nodes?Yes, by carrying or later building the store and transmitter; the network grows with exploration.

I want to be clear about how this work came about.

I did not create every component of this concept from scratch, nor would I claim to have done so. The underlying technologies, research, ideas and discoveries come from many people and organisations who have contributed to human knowledge over time.

What I did was recognise a way those existing pieces could potentially be connected into something larger. I developed the concept in my own mind, worked through how the different elements might fit together, and used artificial intelligence as a tool to help research, organise, articulate and develop the work.

In that sense, the work is collaborative by its very nature. Humanity created the underlying body of knowledge. AI helped me navigate and assemble parts of it. I provided the conceptual direction and kept asking the question: what could happen if these things were brought together?

Innovation has never been exclusively about inventing every individual component. Sometimes the important contribution is recognising a connection that has not previously been made, developing it into a coherent concept, and putting it into the public domain so that other people can examine it, challenge it, improve it or build upon it.

That is what I have attempted to do here.

I am not presenting myself as the sole inventor of humanity’s accumulated knowledge. I am presenting a concept that emerged from connecting that knowledge in a particular way.

And I am making it public because I believe useful ideas should have the opportunity to be explored beyond the boundaries of governments, corporations and large institutions.

Humanity built the pieces. I connected some of them and asked what we could build together.

I made this video partly because I want to challenge the idea that an unemployed person is somehow worthless or useless.

Unemployment and human value are not the same thing.

Our nations contain millions of people with skills, experience, creativity, knowledge, curiosity and ideas who are currently outside conventional employment. That doesn’t mean they have nothing to contribute. It may mean we simply aren’t giving them enough meaningful opportunities to contribute.

I am one example. I didn’t create every piece of technology behind my work. Humanity built those foundations. I connected ideas, developed a concept and used AI as a tool to help bring it together.

The point isn’t that unemployment is inherently worthless.

The point is that people are valuable whether or not they currently have a job.

By dave