RESEARCH RELEASE 0.2 / 20 OCTOBER 2026
Optical benchmark first. Alpha 1 hardware on hold. No physical coupling, apparatus performance or propulsion result is established. The next milestone is a qualified measurement of known photon momentum.
A conditional field, not a discovery
The proposed A1-Y potential obeys ∇²φ − c−2∂t²φ − λ−2φ = 4πGρq, with A1 = −∇φ and F = −mqt∇φ. The effective coupling q and any means of changing it remain physically unspecified. Let qt = q0 + s q1, s = ±1; the source/state-odd amplitude is bK(r), b = qsq1.
A kernel is not an allowed operating point
For nonoverlapping homogeneous spheres, K(r) = GMm f(Rs/λ) f(Rt/λ) e−r/λ(1+r/λ)/r², with f(x) = 3(x cosh x − sinh x)/x³. Using 10 kg, 0.1 kg, density 19,300 kg/m³, λ = 0.10 m and r = 0.20 m gives K = 0.6952 nN. Unit coupling is a normalization, not a fitted, observed or necessarily permitted value.
The historical constraint is already a serious challenge
Murata and Tanaka’s 2015 review summarizes a conventional Yukawa α limit of order 10−3 at λ = 0.10 m from Hoskins and colleagues. If that same coupling applies to this geometry, the kernel gives about 0.695 pN. The v0.1 assumed 95% coupling half-width of 0.234 corresponds to 0.163 nN at 0.20 m—about 234 times larger. This historical comparison alone makes a unit-coupling demonstration a poor hardware target; it is not a newly computed exclusion curve or the latest best limit.
Research basis: Jiro Murata, 2015.
Why α cannot simply be renamed b
Published constraints must be mapped through composition, screening, source geometry, interaction range and the populated material states. A static universal α is not automatically the switchable product qsq1. In fact, a state-independent interaction can cancel from the double-odd observable. Equivalence-principle η is also not b. The superconducting-niobium/copper comparison tests specified materials and its Cooper-pair interpretation has a different bound. Unspecified coupling is a missing model, not evidence that constraints can be evaded.
Research basis: M. P. Ross et al., 2024.
| Gate | Advance only with |
|---|---|
| Physical mechanism | A microscopic interaction and actuator model defining q(state), dimensions, energy cost and stability. |
| Constraint mapping | Reproduce relevant likelihoods or justified published limits using the same composition, geometry and observable; compare current literature before hardware. |
| Detectable allowed prediction | A surviving parameter region whose observable exceeds qualified uncertainty, with conventional alternatives specified. |
| Momentum closure | The external source receives recoil; co-mounting source and target creates internal forces. Account for fields, supports and outgoing radiation. |
| Independent review | A named qualified reviewer signs the model and protocol before apparatus commitments. No such review is claimed. |
What would rule out a specified Alpha 1 version?
Reject the version if its required coupling lies in an excluded region after a valid mapping, if it predicts closed-system acceleration without an external momentum flux, or if a qualified preregistered test rules out its specified force/scaling within the declared domain. A balance that has not passed its controls cannot establish an exclusion. A model with no definite nonzero prediction cannot be falsified by one null result; narrow it before testing.
Equilibrium boundaries are not a loophole
Nonreciprocal materials can change optical and Casimir interactions. Equilibrium analyses still derive forces from a free energy under their stated conditions. A driven thermal system must identify its reservoirs, power supply and momentum exchange. Geometry alone does not establish reactionless motion.
Research basis: David Gelbwaser-Klimovsky, 2022.