MF-01A / Planned campaign
Recover a known force
Establish what the complete instrument can measure before testing a candidate propulsion mechanism.
Required inputs
- A declared force range, payload, measurement bandwidth, averaging interval and maximum acceptable uncertainty, selected from the candidate prediction.
- A calibrated force reference with its uncertainty and a documented relationship between sensor output and force at the candidate's application point.
- A force-balance design, readout, environment and data acquisition suitable for those requirements. Vacuum, cryogenics and interferometry are choices to justify, not automatic requirements.
Controls and measurements
- Calibrate the installed apparatus across the intended range before and after each campaign; include alignment, lever arms, hysteresis and drift.
- Run zero-input records and repeated reference loads at varied levels and run order. Keep raw records as well as processed results.
- Recover an independently predicted optical or electrostatic force and account for uncertainty in beam interception, reflectance or calibration geometry.
- Check whether the proposed averaging time improves precision or instead integrates drift; report bandwidth and environmental conditions with the result.
Reference forces are recovered within a prospectively specified combined uncertainty, and repeatability and drift meet the declared candidate-test requirements. A sensor's advertised resolution alone cannot pass this gate.
Output
An apparatus qualification report: calibration curves, noise and drift records, uncertainty budget, validated analysis and stated operating envelope.
Research basis: Paul Williams, 2021 · Jon R. Pratt, 2006.
MF-01B / Planned campaign
Learn the false signals
Measure how ordinary heat, fields, gas and mechanical connections can imitate the candidate's predicted response.
Required inputs
- The candidate's mass distribution, electrical routing, expected thermal pattern, operating sequence and supporting structure.
- Temperature, pressure and environmental-field measurements appropriate to the selected design, synchronized with force readout.
- Dummy loads and controlled disturbances that reproduce the main non-candidate interactions.
Controls and measurements
- Use matched heating controls: match the spatial and time-dependent thermal response as far as practical, and record residual mismatch. Equal electrical watts alone do not establish equal thermal behavior.
- Characterize thermal expansion, suspension drift and directional thermal radiation separately. Radiation recoil is real force but can be a conventional explanation.
- Map magnetic fields and current paths, including pumps, gauges, calibrators, magnets and return wiring. Account for reaction on external field sources rather than attributing every Lorentz force to a new mechanism.
- Test cable stiffness, feedthrough coupling and electrical pickup; any onboard power option must include its battery, electronics, heating and mass distribution in the supported assembly.
- Where relevant, vary pressure and document outgassing, electrostatic charging and vibration. Orientation reversal alone is not sufficient because ordinary artifacts can also reverse.
Known disturbances are eliminated, bounded or modeled with their uncertainties included. The remaining uncertainty must still resolve the prospectively specified candidate effect. Unexplained control responses pause candidate interpretation.
Output
A disturbance-response map and control matrix that specify which alternative explanations the apparatus can distinguish.
Research basis: Slava G. Turyshev, 2012.
MF-01C / Planned campaign
Test one defined prediction
Determine whether a specified candidate is supported, constrained or unresolved under a declared test protocol.
Required inputs
- A frozen model with predicted force, sign, parameter scaling, operating conditions and a result that would count against it.
- A passed apparatus qualification and disturbance review at the conditions of the candidate test.
- A written run plan, uncertainty calculation, sample-size rationale, analysis code and exclusions decided before decisive data collection.
Controls and measurements
- Interleave candidate-on, candidate-off and matched dummy runs in randomized order. Blind condition labels during analysis where practical.
- Measure a predicted parameter dependence and predicted null configuration, not only a difference between power on and off.
- For near-field candidates, account for both interacting bodies, their supports, field sources and thermal reservoirs, plus momentum in escaping radiation. A force on one plate is not the thrust of the complete assembly.
- Repeat across days and reassembly. Test sensitivity to justified alternative drift and uncertainty models without selecting the analysis that gives the largest effect.
- Package drawings, materials, calibration, raw data, software and known failure modes for independent reproduction. Peer review and replication are separate forms of scrutiny.
A candidate advances only when its predicted response is resolved within the declared uncertainty, withstands the specified controls and repeats. A null result becomes a quantitative bound for the tested conditions; an unresolved signal triggers further controls rather than a propulsion claim.
Output
A reproducible result package with the measured effect or bound, limitations, conventional explanations and the next decision. Independent replication is a subsequent gate.
Research basis: David Gelbwaser-Klimovsky, 2021 · Laila Henkes, 2025.