D02 independent analytic and trace review

The independent audit passed all 144 row inputs, 288 assigned strikers, 1,152
source/profile launch screens, and 36 histories from 18 full traces. No new
trajectory was integrated. Audit runtime was 2.878911 seconds. A separate
60-digit cantilever basis and compensated trapezoidal summation were used;
neither the D02 calculator nor the R07 engine is imported by the reviewer.
Independent means a separate computation inside this project, not an outside
laboratory or an external physical test.

Six focused reviewer tests passed: exact launch-work boundary; dimensional
scaling; rejection of invalid parameters; exact linear-force quadrature;
reconstruction of original full traces for all three contact laws; and
rejection of missing or nonfinite trace samples. The source and input hashes,
output inventory and summary extrema were independently checked.

What the launch calculation says

The fixed protocol assigns each 10 g free striker 5 mJ, corresponding to
1 m/s speed and 0.01 N s launch impulse. The adaptive protocol assigns each
striker 4.95 mJ; its 0.1 mJ event-level probe charge is separate.

Hypothetical profile             Outcome across 288 assigned strikers
0.1 N net force over 5 mm        Fails necessary ideal work bound
0.5 N net force over 10 mm       Passes necessary ideal work bound
2 N net force over 5 mm          Passes necessary ideal work bound
5 N net force over 2 mm          Passes necessary ideal work bound

For the fixed protocol, constant net force over the full 10 mm stroke would
be 0.5 N for 20 ms, reaching 1 m/s. That does not conflict with a much larger,
shorter collision force after release. Fcap*s >= E and Fcap*t >= m*v are
necessary bounds on a separate ideal launch, not evidence that a real actuator
can supply the required force throughout its stroke at the required speed.
The profiles are assumptions, not catalog products or costed assemblies.

What the saved histories add

Among the 36 retained striker histories, the largest sampled motion span over
0 to 0.17 s is 72.069630 mm. Largest forward excursion is 37.496063 mm, largest
backward excursion is 71.718545 mm, and peak rebound speed is 0.426058 m/s.
Largest sampled contact force is 25.093330 N. These extrema can belong to
different records; the summary identifies the exact run and source for each.
The population is the 18 selected full traces, not all 144 row trajectories.

All 36 strikers are still moving at the saved endpoint. Consequently the
finite-window span is not a complete guide/catcher clearance or reset-stroke
requirement. No catcher, braking, end stop or return was modeled. Launch
stroke, free-flight/rebound travel, contact indentation and local beam
displacement are separate quantities.

The maximum trapezoidal work discrepancy is 8.213928550477798e-6 of the initial
main-event energy, below the frozen 0.001 threshold. Maximum absolute impulse
discrepancy is 7.41337117405455e-8 N s; every source meets its separately
computed frozen relative/absolute impulse tolerance. No thresholds were
retuned after seeing results.

Manufacturer evidence

MANUFACTURER-EVIDENCE.json and .txt explain sine/continuous/intermittent/shock
force distinctions, peak-to-peak stroke, load-dependent frequency, attached
moving mass, piezo working curves, preload and mounting constraints. They do
not select a device or establish under-$1,000 performance. Source documents
remain private verification copies and are not included in the public bundle.

Reproduce

python test_independent_mechanics.py --r07-root PATH_TO_R07_EFFICIENCY -v
python verify_device_demand.py --r07-root PATH_TO_R07_EFFICIENCY --d02-source PATH_TO_D02_SOURCE --d02-output PATH_TO_D02_RESULTS --output NEW_AUDIT.json

NumPy and the complete frozen R07/D02 evidence trees are required. The reviewer
refuses to replace an existing audit. No hardware control, source changes,
publication, fracture validation or actuator certification is performed.

D02 summary SHA256:
736ca07de1b279d773c4a1bdc22497274ba777c7030edba5d6fb523dc3159ecd
