The objective
Establish a trustworthy force measurement over a declared range. Then compare a fully specified candidate device with a matched conventional model and control assembly. The first apparatus choice depends on the prediction; there is no evidence-based reason yet to prescribe a terahertz controller, fractal hull or biological power plant.
Three separate campaigns
A / Instrument qualification
Select the geometry and force range from the model. Choose traceable calibration inputs, measure drift and frequency response, and prepare an uncertainty budget. Run known applied forces and blind analysis checks before mounting the candidate apparatus.
B / Known-physics benchmark
Reproduce an established force suitable for the instrument. Optical radiation pressure or a characterized electrostatic calibration can provide a positive control. These are possible routes; final power, enclosure and instrumentation choices require specialist design and safety review.
C / Candidate comparison
Test the proposed interaction only after its prediction is frozen. Use randomized condition order, blinded analysis where practical, matched dissipation and enough independent runs to resolve the preregistered effect. Keep null results and excluded runs with reasons.
Research basis: James E. Polk, 2017 · Paul A. Williams, 2013 · Paul Williams, 2021 · Jon R. Pratt, 2006.
Minimum control matrix
| Possible explanation | Control or diagnostic | Record |
|---|---|---|
| Thermal expansion / drift | Matched heater, settling study, temperature sensors | Force and temperature time histories |
| Magnetic interaction | Current-path changes, orientation and shielding studies | Current, ambient field and geometry |
| Cable or feedthrough forces | Alternative power/data arrangements; measured harness stiffness | Mounting, routing and calibration |
| Outgassing / gas forces | Pressure sweeps, vacuum characterization and matched controls | Pressure, materials and elapsed time |
| Vibration / ground tilt | Seismic/tilt logging, apparatus transfer function | Time-synchronized environment channels |
| Radiation recoil | Optical/RF leakage and directional heat estimate | Power flow and radiation geometry |
| Analysis bias | Randomized labels and locked processing | Raw data, code version and exclusions |
Decision rules
Before collecting candidate data, choose the minimum scientifically relevant effect, confidence procedure, independent-run design and allowed exclusions. An apparent signal advances only if it survives the apparatus-specific uncertainty budget and intended controls, with a physically consistent sign and scaling.
A null result establishes an upper bound only within the measured setup and parameter range. It does not disprove every possible interaction. A positive result earns a replication campaign; it does not immediately earn a vehicle claim.
Research basis: M. Tajmar, 2022 · O. Neunzig, 2021 · Slava G. Turyshev, 2012.
Required output package
- Versioned model, drawings, material specification and complete system boundary.
- Calibration history, drift study and combined uncertainty budget.
- Raw synchronized channels, run order, environmental records and exclusions.
- Frozen analysis and comparison against alternative explanations.
- Replication instructions, a conclusion limited to the evidence, and the next decision.