Status: design note, hypothesis level. The mechanism is now specified and the estimates are quantified; nothing here is a measurement.
Supersedes the material strategy in the-star-forge.md and star-resonator-build.md.
The mechanical and geometric reasoning in both still stands; what changed is that no
single material is asked to do two incompatible jobs, and the target of the device is
now stated as a flow rather than a reserve.
Earlier generations built the pentagonal frame as a continuous soft ferromagnetic body — iron-silicon with a cerium oxide sintering aid — and wound coils at the vertices. The reasoning there was sound (one continuous magnetic circuit, no air gaps, high permeability, low hysteresis, adequate resistivity). The problem is that the same body was also expected to supply five-fold geometry, and a soft magnetic material cannot: softness and high-frequency resonance are opposed, and no ferromagnetic alloy has quasicrystalline structure.
This note splits the functions:
It also replaces the earlier framing. The device is not a tap on a stored reserve. It is a turbine in a flow, and the flow is what the rest of this document is about.
The hypothesis holds that space itself has a tendency to subdivide — to grow, to generate new volume — and that this tendency is active at the smallest scales. Left unopposed it would make space expand continuously in the immediate neighbourhood of matter.
Matter is coherent, so the subdivision is opposed locally. The counterforce is a condensate: a paired, neutral, coherent medium that associates with matter in proportion to its structure, and whose tension holds the local volume together.
The proportion is the interesting part. Baryonic structure offers one site per proton for an electron pair; if each such site associates with a condensate pair, then the amount of condensed medium is proportional to the visible mass. That proportionality is not decorative — it is what makes the medium's response track the baryonic distribution, and it is the property the galactic data actually constrains.
Perturb the condensate and it restores. In the linear regime the recoil is proportional to the displacement, the work done equals the work returned, and the cycle closes on itself. Any experiment that stays in this regime will net zero, regardless of geometry, resonance or material quality. This is the single most important practical statement in the document: a null result inside the elastic regime is not evidence against the hypothesis, it is evidence that the wall was not reached.
Push far enough and the medium can no longer deform — it must convert. In phase-transition language this is the spinodal, the limit of metastability; in condensate language it is the critical field, where pairs break rather than strain. Same wall, two vocabularies.
The observable signature is hysteresis. Linear recoil returns along the path it went out on. A spinodal crossing does not: it leaves a loop, and beyond it the medium does not fully restore. So an amplitude sweep is the primary diagnostic — drive harder, watch the response, look for the transition from proportional to looped.
Past the wall, pairs convert. If the conversion is annihilation, the mass leaves the condensate's books permanently, because annihilation is not reversible: photons are not reabsorbed into a pair at any useful rate. That irreversibility is what makes the converted energy available rather than merely lent.
The depleted region does not stay depleted. The surrounding medium is under tension, so it closes inward, exactly as a bag re-homes toward the common centre of a mass distribution. That inflow is the power source. Not the reserve — the flow.
This also explains two long-standing observations without additional machinery: a cluster having a single gravitational centre while its contents retain local centres is coarsening, the same behaviour as droplets merging to minimise surface area; and the force law that results from a tension-bounded cloud is 1/R, which is precisely what flat rotation curves require.
The device is a turbine in the dark matter wind. That classification is what makes the idea tractable: turbines are bounded by the flow, not by the fuel, and their efficiency has a calculable ceiling that can be derived before anything is built.
None of these is the hypothesis, but each is a published model that arrives at one of its structural features from an independent direction. The convergence is the reason the idea is worth testing rather than dismissing.
| Established model | What it converges on |
|---|---|
| Two-fluid / Ginzburg–Landau | A condensed and a normal component in equilibrium, with a surface energy term. This is the formalism the mechanism above is written in. |
| Verlinde, emergent gravity (2016) | The dark component as an elastic response of dark energy to baryonic mass; apparent dark mass proportional to baryonic mass. |
| MOND / radial acceleration relation (McGaugh 2016) | Galactic dynamics determined entirely by the baryonic distribution, with almost no scatter — proportionality, measured. Baryonic Tully–Fisher: M_bary ∝ v⁴. |
| Superfluid dark matter (Berezhiani & Khoury 2015) | A condensate with zero viscosity, giving collisionless halo behaviour and MOND-like galactic dynamics. |
| Asymmetric / mirror dark matter | Why ΩDM/Ωb ≈ 5.4 rather than 10⁻⁶: a shared asymmetry produces a ratio of order unity. |
| Dissipative dark matter (Fan, Katz, Randall, Reece 2013) | A hidden sector with its own gauge force: "conductive", but only for dark photons. |
| Casimir effect | A field doing measurable work at electron-volt energies from boundary geometry alone, with no particle-scale wavelength anywhere. |
| Haloscopes (ADMX and successors) | Resonant detection of a coherent dark field at f = mc²/h, with signal power ~10⁻²² W and Q ~ 10⁵. |
The empirical weight sits with the radial acceleration relation: the proportionality the hypothesis requires is the proportionality the data shows, at galaxy scale, with scatter small enough that it is currently the strongest argument against particle dark matter in that regime.
Mass-energy conversion is 25 GWh per gram. Equivalently:
About 10 MWh a year: 0.4 mg of mass equivalent. The local dark density supplies that from a cube roughly 90 km on a side. Availability is therefore not a constraint at any scale of interest — this is the part of the thesis that holds unconditionally. The fuel is coextensive with space, and everyone's space is enough.
Local dark density 0.3 GeV/cm³, streaming at ~300 km/s, gives a mass flux of about 9×10⁶ GeV cm⁻² s⁻¹. Across one square metre, at perfect conversion with no losses, that is ≈14 W. That is the absolute ceiling for a given aperture, and it is the Betz limit of this design: not a materials problem, a flow problem.
The ceiling assumes total absorption. The same property that makes the medium dark — negligible coupling to ordinary matter — sets the absorption cross-section. Realistic capture fractions are many orders below unity, because any coupling large enough to capture efficiently would already have appeared in every experiment ever run.
This is the number to determine. Everything else is engineering around it.
The wall's height is the condensation energy density of a paired fluid, which real
superconductors measure directly. Critical fields of 0.05–1 tesla correspond to
µ₀H²/2 of roughly 10³ to 4×10⁵ J/m³.
Per litre of active volume, fully driven past the wall, that is **order joules to hundreds of joules per conversion**. That is the per-cycle ceiling, and it comes from measured condensate physics rather than from the hypothesis.
Power is the product of yield and rate:
consume it first
The tension inside the design: at the critical field the losses are also maximal, so the high-rate regime is exactly where the drive cost rises fastest. The honest expectation is that the useful band sits above the spinodal and well below the point where the medium is destroyed rather than converted.
If the device produces power P for time t, the mass equivalent is
Δm(kg) ≈ 1.1×10⁻¹⁷ · P(W) · t(s)
So 1 kW for a day is about 1 µg; 1 mg requires about 25 MWh, which is a power-station experiment rather than a bench one. Arriving from the other direction, consuming an entire cubic metre of local dark matter is about 0.5 nanogram. The instrumentation requirement is therefore nanogram-to-microgram resolution, and the run must be either high-power or long.
Drive amplitude swept upward, response recorded, hysteresis looked for. Proportional and reversible below the wall; looped and irreversible above. This requires no exotic apparatus and it is the test that distinguishes "the wall was not reached" from "the wall is not there".
Two matched samples in one chamber, one driven and one not; measure the difference. This defeats the artefacts that share the signal's signature — electrostatic force scaling with voltage, ion wind, magnetic attraction from the core and coils, Joule heating driving convection, thermal expansion of the mount, and field-driven adsorption. Required: full Faraday enclosure with a filtered feedthrough, thermal isolation with steady-state waits, a non-magnetic balance, and someone else throwing the switch.
Note the trap: **a voltage-proportional weight change is exactly what electrostatics produces.** Without the enclosure and the matched reference, no result from this test means anything.
rate overproduces the galactic 511 keV line measured by INTEGRAL/SPI since 2003, which is already fully attributed to conventional positron sources. This bounds the electron identification specifically, and it is existing data.
recombination. The hardest constraint on any purely baryon-coupled scheme.
provides under the 1/R treatment.
lags. Natural for a collisionless component; awkward for a medium, and the sharpest disagreement to answer.
Scan rate is the bottleneck in coherent dark-field searches, limited by Q times tuning range. A quasiperiodic cavity does not have one resonance — it has a self-similar ladder of them at φ-scaled frequencies, present simultaneously. A Fibonacci/five-fold cavity could therefore search several mass windows at once. That is an unaddressed problem in a live research programme, and it uses the geometry already built, independently of whether the energy hypothesis is correct.
Snoek's limit: (µ_r − 1) · f_r ≈ const ∝ M_s. Permeability and resonance
frequency trade inversely. µr of 3000 lives at kHz–MHz; µr of 1–10 is what resonates
at gigahertz. The escape is anisotropy shape: for easy-plane anisotropy the limit
relaxes to µ·f², which permits usable permeability at much higher frequency. That is
the basis of planar hexaferrites (Co₂Z and relatives) and the reason hexagonal ferrites
are what millimetre-wave circulators are made of.
Low hysteresis and high frequency are opposed for the same reason seen twice: low hysteresis means low anisotropy, low anisotropy means low resonance frequency. The material candidate is barium or strontium hexaferrite, or — unresolved — an engineered element geometry supplying artificial permeability where no crystal can.
An Al–Cu–Fe icosahedral quasicrystal as a thin film on the fired core. Its magnetic
contribution is nil and should not be claimed: the i-phase is weakly diamagnetic to
weakly paramagnetic, µ_r ≈ 1. It is chosen for geometry and it delivers geometry.
What it contributes:
a lossy conductor is how unwanted modes are damped.
makes the coating a surface impedance sheet; super-skin makes it a shield.
tiling is a quasiperiodic frequency-selective surface: self-similar multi-band response, no rational coincidences between resonances.
Thermal compatibility works: coatings run roughly 7–11 ppm/K, ferrites 8–10, and the process temperatures stack correctly — fire the core at 1000–1300 °C, coat and anneal at 600–800 °C.
Fibonacci-proportioned elements (1, 2, 3, 5, 8) at the bridge points of the upper and lower pentagons, coupling the arms and setting the operating band.
φ is the Hurwitz extremal ratio — slowest-converging continued fraction, the worst approximated by rationals. Under KAM theory that makes φ-ratio coupling the most robust against mode locking, which is the technical content of "the right angular ratios for stability". A Fibonacci ladder is a quasiperiodic network with measured self-similar transmission spectra and Cantor-like band structure.
Above roughly a gigahertz, discrete capacitors become unusable because wiring parasitics (1–5 pF) are the same order as the required element; the network becomes distributed stub lengths in Fibonacci proportion.
closure of a two-strand braid with five crossings, ten nodes.
bias selects which handedness of circular polarisation the structure couples to. Reversing bias flips the selection.
forbids five-fold rotation in a periodic lattice. A decagonal quasicrystal is a cut-and-project from a four-dimensional lattice plus one periodic stacking axis.
A condensate is a field-theoretic object. It responds to boundary geometry and coherence, not to single-particle collision energy — the reason Casimir works at electron-volt energies with millimetre plates.
So frequency does not set spatial reach: 1 GHz carries 4 µeV per quantum, while the electron Compton scale is 511 keV, a factor of 10¹¹. What sets reach is:
l_B = √(ħ/eB) — 26 nm at 1 T, 8 nmat 10 T.
d, via momentum ħ·2π/d — keV-scale at d = 1 nm.The two variables worth engineering are therefore bias field magnitude and coating pattern period. The φ network sets coherence, not reach.
incident flux converts. It is bounded above by the requirement that the medium remain dark, and no design work matters until it is estimated or measured.
scales toward particle dimensions, and it is the least developed part of the design.
microwave surface. Sputtering gives the right film but needs a target and a post-anneal with tight composition control.
all, or a structured element array, is unresolved.
through a slow peritectic; slow cooling gives the equilibrium assemblage, not the quasicrystal. A powder-in-tube approach — phase made before it enters the vessel — avoids the problem entirely.
mechanism requires the conversion branch to be genuinely irreversible and the refill flow to be the delivered power. Any experiment that cannot distinguish those two will produce an ambiguous result.
If the conversion branch is real, the consequences are large. The medium would be identified, its distribution shown to be coextensive with baryonic structure rather than concentrated in distant objects, and a direct conversion path established from that medium to electrical energy. Because the fuel is proportional to visible matter and visible matter is everywhere, no one would be short of it — the quantity required for a household's annual electricity is sub-milligram.
All of that is conditional on results that do not yet exist. The value of the document at this stage is that the claim is now small, specific, and falsifiable: one cross-section, one hysteresis measurement, one mass difference.
The chirality result needs no core, no ferrite and no microwave apparatus: grow a mixed alum co-crystal (potassium aluminium sulfate with iron and chromium substituted) in a vessel of the required geometry, under DC bias, and check whether domain handedness follows bias direction. Hydrated, paramagnetic, isomorphous, growable in a jar with an oil bath — a couple of hundred dollars rather than a couple of thousand, and it tests the one claim that survives independently of the energy question.