Correa (PAGD / "Aether Motor"): Science Contrarian review

Reviewer role: Science Contrarian (judges by experiment, not consensus; applies the same rigour to the inventors as to their critics).

Source gap (repeated from Step 1, verified)

Only the rexresearch.com printout (44 pp., full text layer) is readable. The two patent PDFs are image-only scans; their body text was taken from the same patents reproduced in the printout. Not obtained: raw measurement records, the full Labofex Development Report S3-001, the 1996 Denver symposium paper, any Aether Motor documentation, any independent test report. Every number below is the inventors' own. Where a conclusion rests on [memory] alone, it is marked.

Source tags: [read] = in the printout (I extracted the text with pdftotext and re-read Table 7's page as an image, p.25); [web] = URL given; [memory] = with confidence; [calc] = my own arithmetic on [read] numbers.

Corrections to Step 1

1. Step 1 cites only Table 7 (218-563%) and Table 14 as the energy evidence. The compilation reports four separate accounting methods, and they disagree with each other. [read] Table 5 (open-circuit voltage to "capacity" conversion) gives up to 776%, and 3.9x in the worked examples; Table 6 (resistive discharge integration) halves the loss estimate; Table 7 (instantaneous DC V and A) gives 218-563%; Table 8 (pre/post resistive discharge curves) gives 406% to 3485%. The authors themselves say Table 5 "in all likelihood, overestimates the actual values" and that the Table 7 method "ignored" battery time of delivery. Step 1's summary "computed from mean DC volts and amps ... authors admitting battery state/time was ignored" is right but incomplete.
2. Step 1 states the worked example as "5.8e-8 g of aluminium". The text gives a cathode density of 1.86 g/cm3, which is not aluminium (about 2.70 g/cm3 [memory, high]). The mass 5.8e-8 g equals 3.1e-8 cm3 x 1.86 [calc]. Minor, but a data-quality flag Step 1 missed.
3. Step 1 says the 0.5 MeV ion energy is "computed". It is more specific than that: it is derived by dividing the measured pulse energy by the number of ions, so it is not an independent measurement (see Claim 2).
4. Step 1 is otherwise accurate: gain figures, dyne values, patent numbers and abstracts match what I read. No Step 1 claim of independent testing was made, and I found none either.

Claim 1: Large-area, large-gap cold-cathode tube self-pulses at much lower fields/current densities than Fowler-Nordheim (FN) predicts

a. Maths: FN is a tunnelling law for a clean, smooth surface, J ~ F^2 exp(-b/F). The compilation compares the macroscopic field (V/gap) to FN. No local field is measured. [read] Table 7 gives "extinction potential" 22-81 V/cm; that is a gap-averaged figure. Internally nothing contradicts itself; the comparison is just not like for like. b. Experiments: Field emission at macroscopic fields far below the FN smooth-surface value is routine, because microprotrusions, oxide films and adsorbates enhance the local field by factors of 10-1000 (the "beta factor") [memory, high]. Relaxation oscillation in a negative-resistance discharge with a capacitor and ballast is also textbook behaviour [memory, high]. Neither is excluded by the inventors' data. c. Tested? Only by the inventors. No independent replication that I could find [web: search of Correa/PAGD, results are almost all the Correas' own pages and patents, e.g. https://www.aetherometry.com/Labofex_Plasma_Physics/Archive/PressRelease.html]. I could not read the one 2007 third-party PDF found (http://www.rohnermachine.com/Files/CPAGD.pdf); the fetch returned no text, so I make no claim about it. d. If true: either FN with field enhancement fails or a new emission trigger exists. The mainstream explanation would need only surface physics; nothing fundamental changes.

Strongest honest case: the self-pulsing regime in a large-gap, low-pressure (0.2-1.5 Torr [read]) tube with 14 reported configurations is a legitimate discharge regime, the pulsing rate varied systematically with pressure, plate work function and ballast, and "break in" over more than 2x10^6 pulses was reported. That is an orderly dataset, not noise. Against: no local-field measurement, no surface characterisation, so "violates FN" is unproven.

Verdict (split): self-pulsing behaviour: Plausible, untested (independently). "Violates FN": Never properly tested (not dismissed without testing; simply never examined by anyone else that I can find).

Claim 2: Ions at ~0.5 MeV and reaction forces of 100-300 dynes, 100x vacuum arcs, as per Aspden's Law

a. Maths: I rechecked the worked example. 5.8e-8 g / 26.98 u = 1.3e15 ions, correct. 4.8e5 eV x 1.3e15 x 1.6e-19 J/eV = about 100 J [calc], close to the stated net pulse energy of 86.4 J, so the "ion energy" is simply the pulse energy divided by the ion count. That makes it circular: it assumes all the claimed surplus energy ends up as ion kinetic energy, then uses it as evidence of anomalous energy. The reaction force, 143.6 dyn, equals momentum m*v (about 9.9 g cm/s) divided by about 49 ms [calc]; fine arithmetically, and the 245 dyn "reaction balance" number is the only direct force measurement, with no details given. The k values in Table 15 (about 100 x 10^2... i.e. force/i^2) scale as claimed. The cathode density of 1.86 g/cm3 is wrong for aluminium (flag). b. Experiments: In vacuum arcs, cathode-spot ions carry roughly 20-150 eV [web: https://www.osti.gov/servlets/purl/810482], with the Tanberg/Kobel "anomalous" forces historically explained by plasma pressure and ion jets [memory, medium]. A 560 V supply cannot accelerate an Al ion above about Z x 560 eV (Z up to 3 in arcs; max about 7 keV even for Al13+) by the applied field [calc]; 0.5 MeV would need an acceleration mechanism the paper does not demonstrate. c. Tested? Only by the inventors. No ion energy spectrometry (time-of-flight, retarding-field analyser) is reported in what I read. Dismissed without testing: no evidence either way. d. If true: the Lorentz force law / momentum conservation would need an additional term (Aspden's), and cathode-spot theory would need rebuilding.

Strongest honest case: the Graneau water-arc explosions and Tanberg's 1930 forces are real, and the inventors measured force with a balance, which is a direct and testable technique. A balance measurement of 2.5 mN impulse-type forces from a 1e-8 g ablation event is surprising enough to merit repetition with calibration. Against: ion energy is inferred, not measured; mass per pulse comes from "metallographic data" with the wrong density; the Aspden fit has a free constant (the mass ratio 49,185 is called "in the 10^4 range").

Verdict (split): "0.5 MeV ions": Never properly tested (inferred circularly from the energy claim; strong form unsupported). Force 100-300 dyn: Never properly tested (single reported balance run, no independent repeat). Aspden's law as the explanation: Never properly tested / Not testable as stated (free proportionality).

Claim 3: Double-ported circuit captures more energy (218-563%) than it consumes

a. Maths. I recomputed Table 7 from its own columns [calc, from the image of p.25]. Column 8 "DP Watts" equals (X - Y) x I, not X x I. Example, experiment 1: X = 562 V, Y = 350 V, I = 0.65 A gives (562-350) x 0.65 = 137.8 W, which matches the table; but the drive pack terminal power is X x I = 365 W. The text defines X as "drive pack DC voltage" and I as "drive pack output current". Using X x I, the CP/DP ratio for the 14 rows is: 128, 78, 74, 104, 95, 91, 80, 66, 66, 146, 99, 87, 116, 67 percent. Only 4 of 14 exceed 100%, the median is about 89%, and the "218-563%" collapses to 66-146%. Caveat: I cannot rule out that Y is defined differently from the text (for example, that the ballast drop is a separate loss the authors chose to exclude); but the drive pack still has to supply the current through the ballast, so excluding it is not a fair input measure. Further arithmetic slips: Table 7 row 10 lists 530% where CP 387.3 W / DP 67.2 W = 576%; row 14 lists 228% where 257.7/156.8 = 164%; row 9 lists DP 112.45 W where (X-Y) x I = 129.8 W. Second internal inconsistency: Table 8 group 2 reports the drive pack lost 7.3 Wh over 18 min (about 24 W), whereas Table 7 shows 0.4-0.75 A drawn at about 560 V (about 200-420 W, i.e. about 60-125 Wh in 18 min) [calc]. Table 5's worked example claims 0.988 kWh gained in 21.5 min (2.8 kW average) while the charge pack input in Table 7 is 210-470 W. Those methods cannot all be right; the authors concede the first overestimates. Third: measurement method. [read] "Banks of Beckman rms multimeters (30 and 100 kHz bandwidths)" were used "for all current measurements" of a highly pulsed waveform (pulses as brief as ns-scale, 150 A peaks). A battery gains energy from the mean current x voltage; an rms reading of a pulsed current is larger than the mean. Whether this inflates CP power depends on meter mode, which the text does not state. Treat as an unexcluded artifact, not as proven. Fourth: energy-per-pulse (Table 14, "reactor" 25 to >400 J from 5-15 J input) comes from oscillographic AC readings across capacitors (5.3 uF or 17.5 mF series). The energy stored in 17.5 mF at 560 V is 0.5 x 0.0175 x 560^2 = 2.7 kJ [calc]; pulses of 100-350 J are a fraction of what the bank holds. Integrating v x i across an AC-coupled capacitor gives circulating (reactive) energy that returns each cycle; it is not net energy. "Input energy per pulse" is evidently the drive current during the pulse, not the average draw between pulses. The inventors do not show the integration windows. b. Experiments: Conservation of energy is among the most heavily tested laws; but a measurement error is far more probable than a new source. Battery-based accounting is a known weak method: lead-acid open-circuit voltage relaxes for tens of minutes and "surface charge" misleads capacity estimates [memory, high]; the packs had 12 months of ageing and 46 vs 29 cells with different loads. The paper's own admission that "absolute charge capacity is itself a variable" undermines Tables 5 and 8. c. Tested? Only by the inventors. Not independently replicated [web: no independent report found]. Was it dismissed without testing? I found no record either of a neutral lab testing it or of one being refused; Axelrad's note says the lab was closing for lack of funds [read], so a funding problem is documented but not a suppression story. d. If true: energy conservation would need an unmetered source (zero-point energy); thermodynamics and QED would need revision.

Strongest honest case: the authors used four methods, including pre/post resistive discharge curves that avoid instantaneous meter readings, and reported consistent sign (gain greater than loss) across them; they published their cell counts, battery models and uncertainty notes in a patent, which is more openness than most over-unity claimants. A decisive test is easy and cheap: run the circuit from a calibrated DC supply with a calibrated power analyser (true mean power at both ports, wide bandwidth) and a resistive/calorimetric load instead of a battery. They never did this, to my reading. Against: the Table 7 headline numbers do not survive recomputation from their own columns; the methods disagree with each other by factors of 2 to 10; accounting is by battery capacity change of a few percent; everything is by the inventors with no blinding or independent instrumentation.

Verdict (split): Table 7 "218-563%" as stated: Internally inconsistent (drive power appears to omit the ballast drop; arithmetic slips in rows 9, 10, 14). The general claim of net gain: Never properly tested (no neutral, calibrated test on record; not shown to have been refused). I do not call it Tested, not supported, because no independent test exists.

Claim 4: Surplus comes from the vacuum (zero-point / aether energy)

a. Maths: no derivation. The paper offers Aspden's vacuum-energy density and speculation about pair creation and "photonium". The authors write that they do not wish to be bound to a particular theory and that the circuit "clearly draws on a further source of energy input". b. Experiments: no laboratory has extracted net energy from the zero-point field [memory, high]. The Casimir effect is real, but it is a force between boundaries, not a power source. c. Tested? No test of the mechanism exists; it is conditional on Claim 3, which is itself unsupported. d. If true: QED/thermodynamics would need new physics; quantitatively unspecified.

Verdict: Not testable as stated (no mechanism, no prediction other than the gain itself); conditional on Claim 3.

Claim 5: The tube acts as a DC-to-AC inverter driving induction/synchronous motors, speed set by pulse frequency (US 5,416,391)

a. Maths: consistent with a relaxation oscillator with capacitive coupling to a motor winding; synchronous speed = 120 f / poles, and the printout shows linear rpm vs pulses per second [read: FIG. 8 description]. b. Experiments: relaxation oscillators and spark-gap/neon-tube inverters are known [memory, high]. Nothing here conflicts with experiment. c. Tested? Inventors' own motor and torque tests (rope-and-pulley blocked-rotor torque) [read]. Not independently replicated, but there is also nothing extraordinary to replicate. The motor tests report no overall efficiency, so they say nothing about over-unity. d. Nothing in mainstream physics must change.

Verdict: Plausible, untested (independently). This is the least controversial claim; its value is a working device, not a new law.

Claim 6: "Aether Motor" produces electrical power with no input beyond orgone accumulators, insulated bodies or ground (Reich's orgone motor)

a. Maths: none given. b. Experiments: contradicts energy conservation as normally measured; no data are in the files. A ground connection and insulated bodies can pick up atmospheric or ambient electrical energy (the atmosphere has about 100 V/m fair-weather field [memory, high]), so the claim as phrased ("no external input") is not even well defined. c. Tested? Only a personal testimonial by Axelrad, who states he has no plasma-physics training [read]. No measurements, no replication. Not dismissed without testing: simply undocumented. d. If true: conservation of energy would fail in a very large way.

Verdict: Not testable as stated (testimonial only, no specification). Strictly it would be Never properly tested, since it is a testable device in principle, but nothing in the files lets anyone run the test.

Researcher vs critics (rigour on both sides)

• Researcher: the tables mix definitions of "input" and "output"; arithmetic slips are present; the density of "aluminium" is wrong; ion energy is derived circularly; measurements rely on rms meters and battery state-of-charge inference; the surplus-energy explanation is declared unnecessary and then asserted anyway. The press release calls the system "fully protected" by patents; a patent grant is not a validation of the data. These are the same faults the contrarian would object to in a mainstream paper.
• Critics: I found no published, named critique or null replication of PAGD in the searches done, so mainstream rejection looks like neglect rather than a documented failed test. That absence is itself unproven either way, and I will not state that tests were done or failed.

Single decisive experiment

Measure true mean input and output electrical power with a calibrated wideband power analyser (or calorimetry of the whole circuit including the tube, ballast and load), over a long run and with the battery packs replaced by resistors or by an energy-neutral loop; and measure the ion energy directly with a retarding-field analyser. If the output stays below the input at the true measured values, Claims 3-4 fail; if above, there is something that deserves an independent lab.

Summary of verdicts

1. FN / self-pulsing: pulsing Plausible, untested; FN violation Never properly tested.
2. 0.5 MeV ions and 100-300 dyn forces: Never properly tested (ion energy inferred circularly).
3. Net energy gain: Table 7 as stated Internally inconsistent; general claim Never properly tested.
4. Zero-point source: Not testable as stated.
5. DC-to-AC motor inverter: Plausible, untested.
6. Aether Motor: Not testable as stated (testimonial only).