Research note / Cycle 020
A witness with nowhere to stand
A proposed witness placement fails a support-independence walk-through; five numerical cases bound the remaining assumptions.
The third point needs a floor
The Space Elevator Initiative still has two mounted measurement channels and no validated independent witness. Reed and Voss retrace the unloaded support route after the SE-M18 clearance mock-up was removed. The nearest spare stand position rests on the same base as KIS-61 and KB-2. They mark it unsuitable as an independent datum before installing or measuring anything.
Their second route reaches an adjacent bench but brings a cable back through the occupied clamp reach. The third route remains a drawing without a stand or agreed structural support. The route inspection is an authored qualitative performed task: no position reading or bracket-deflection measurement is recorded.
An independent number can be only partly independent
Five separate numerical controls place two target channels and one hypothetical witness at constructed positions. When both targets move from 8 to 12 micrometres, an assumed unmoving witness at 8 shows a +4 micrometre mean discrepancy. If the witness moves with the base, all three show 12 and the discrepancy vanishes. An imperfectly coupled witness at 10 only shows +2. None is a sampled instrument reading.
A route that looks separate on a drawing may still share floor movement or a force path. Dover Motion's discussion of Abbé error concerns an offset between a reading point and point of interest; it supplies neither a measured Meridian reference nor the mechanical stiffness needed to resolve that uncertainty.
An unbuilt answer
The earlier A third point is not always independent set a condition for the third channel. The current walk-through removes a tempting but invalid location from the list. It does not install a witness or reopen strip loading.
Next the team must obtain a support-path drawing with clear load and return paths, qualify any proposed base and quantify the full 1000 micrometre bidirectional grid against the 25 micrometre budget and separate 6 micrometre return allowance. NQ-17 remains returned.
Results at a glance
| Constructed case | Left (µm) | Right (µm) | Witness (µm) | Difference (µm) | Mean minus witness (µm) |
|---|---|---|---|---|---|
| baseline | 8 | 8 | 8 | 0 | 0 |
| separate witness assumed | 12 | 12 | 8 | 0 | 4 |
| same-base witness | 12 | 12 | 12 | 0 | 0 |
| partially coupled witness | 12 | 12 | 10 | 0 | 2 |
| one-sided fault | 12 | 8 | 8 | 4 | 2 |
External research and methods
- Abbé Error · Dover Motion · Motion Control Handbook, Abbé Error section (Undated online handbook; retrieved 2026-10-26). Offset-arm and angular-error explanation.