Concept study / Tests & research methodsConceptual illustration. Shapes, colors, and motion are not measured data.
A force sensor measures its response through a calibration model. It does not directly identify the cause of that response. Small apparent forces may arise from thermal expansion, buoyancy, vibration, electromagnetic interactions, mechanical contact or changes in the sensor itself. NIST’s force-metrology work illustrates why calibration and uncertainty belong together. For a geometric-propulsion hypothesis, the first task is to predict the signal and its ordinary competitors. A propellantless device can still exchange momentum with radiation or an external field; a claim of isolated reactionless thrust is stronger and requires a correspondingly complete account of the system boundary.
Calibrate in the relevant regime
Use a reference appropriate to the expected magnitude, direction and timescale, and verify the readout before and after testing. Check linearity, hysteresis, zero drift and frequency response as relevant. A calibration at a much larger static load does not automatically establish sensitivity to a tiny dynamic effect. Record the complete sensor and readout configuration, including filtering. The proposed analysis should estimate the force from the calibration function and propagate its uncertainty rather than treating a voltage or displacement trace as if its physical conversion were exact.
Build controls for each plausible artifact
Run randomized active, inactive and matched-power conditions with blinded labels where practical. Monitor temperature, vibration and other environmental channels that the mechanism or apparatus makes relevant. Reverse the claimed thrust direction without unnecessarily changing cable routing or support geometry, then repeat with a different arrangement to expose remaining couplings. Controls must be tailored to the apparatus; there is no universal checklist that proves a residual is fundamental. Before interpretation, ask whether a conventional model predicts the signal’s timing, sign and dependence at least as well as the proposed mechanism.
Account for momentum exchange
Define whether the boundary includes supports, power supply, emitted radiation, airflow and external fields. A force on one part of a system can be balanced by another part or by momentum leaving the boundary. For a propulsion claim, measure a mechanically meaningful net effect rather than an internal stress alone. Independent replication should use a different instrument design where possible, because identical setups can share identical artifacts. If no effect is resolved, report an upper bound in the tested conditions and explain which proposed parameter values it excludes.
The mathematical lens
The precise flux expression depends on the chosen boundary and matter or field model. Internal forces alone do not establish isolated net thrust.
Turn it into a question.
- Question
- Does the residual survive calibration, matched-power controls and independent apparatus?
- Method
- Predict artifact signatures, randomize conditions and propagate calibration and environmental uncertainty.
- What would count against it?
- A signal tracking heating, supports, cables or sensor drift does not establish the proposed propulsion mechanism.
A residual becomes informative when ordinary causes have quantitative predictions too.
Follow the original work.
These sources support the mechanisms and limits discussed here. They do not endorse Hill’s framework. Notes distinguish research proposals, reviews, and observations.
Continue the investigation
Edition 02 · 18 October 2026 · AI-assisted synthesis. See editorial method.