Central claims: Haisch-Bernard
Source tags: [read] from the documents; [memory] from training. No verdicts here; the reviewers judge. The patent and the conference abstract are by inventors with a commercial stake (assignee Jovion Corporation); Forward (1984) is an independent published proposal funded by the US Air Force and Hughes.
No source gap: all three documents have readable text (Forward 1984 and the patent are OCR, checked against page images for equations and figures).
- Atoms (preferably noble gases) passing into a Casimir cavity (a gap of about 0.1 micron between conducting surfaces for noble gases; the patent pairs hydrogen with d = 250 angstrom or thereabouts, using the patent's own cutoff convention wavelength = d, which is looser by a factor 2 than the textbook ideal-plate 2d) lose orbital energy, about 1-10 eV per transit, by a "Casimir-Lamb shift", because suppressing vacuum-field modes upsets the balance between Larmor radiation and zero-point absorption that holds the electron in orbit. [read: Patent7379286 p.10-12; Haisch_abstractSSE10 p.1] The 1-10 eV figure is stated as an assumption ("we assume"), and the 0.1 micron cavity is supported in the patent by an uncited statement attributed to Jordan Maclay (the inventors' characterisation). [read: Patent7379286 p.11]
- On leaving the cavity the atom re-absorbs energy from the ambient quantum vacuum field, so repeated cycling delivers usable heat or power drawn from the vacuum, "extracted locally, replenished globally", without violating thermodynamics. [read: Patent7379286 p.12, 19; Haisch_abstractSSE10 p.1] The patent itself concedes that if the energy release and the re-supply are both local, or both global, "no net energy is provided", and proposes an asymmetric entry/exit design (FIG. 8: absorbing glass layer upstream of the conducting cavity, none downstream, conducting length chosen so exiting atoms lack "substantial access" to radiation from the absorber; no length or number given) to make the release local and the re-supply global. It also proposes moving-wall variants (stretched photonic-crystal polymer, MEMS rocking plate; FIGS. 6-7) with no exit-specific mechanism. No independent claim requires net energy gain or recites the asymmetry. No measurement is reported for any embodiment. [read: Patent7379286 p.18-21]
- The underlying physics is stochastic electrodynamics (SED): the hydrogen ground state is a classical orbit at the Bohr radius sustained by Larmor emission balanced by zero-point-field absorption (Boyer 1975, Puthoff 1987, Cole and Zou 2003-2004), and the same balance is asserted, though not proven, for all atoms (the patent says so itself), and "believed" to underlie molecular bonds. [read: Patent7379286 p.12] [read: Patent7379286 p.10-12]
- The patent's own estimates give net power: 21-210 W (channels, 10x10 cm plates), 2150-21500 W (1 cm^3 tunnel block) and 1-10 W (1 sq. cm bonded-wafer device at 8 atm), with pumping power of about 1.7 mW. These are estimates only; no measured output is reported. The printed pumping figures contain an internal inconsistency (1700 Pa versus 17 atmospheres, about 1000x apart). [read: Patent7379286 p.13-15, page images]
- Forward (1984): a stack of charged ultrathin aluminium foil leaves, held apart electrostatically just below the Casimir force, can yield electrical energy as it collapses under the Casimir force and can be recharged. He treats the vacuum-fluctuation field as probably conservative (hydroelectric-dam analogy), i.e. a battery, and notes that no good metal-plate Casimir measurements existed at 1-20 nm. [read: Forward1984 p.1-3]