part-specs: auto-empty dock fan — an airflow number for the dock, and a check on the "weak-ish" note

[serengon]

Following the submit instructions in contributions/part-specs/README.md, announcing a spec sheet I just opened as #64.

part-specs: auto-empty dock fan — airflow requirement + candidate specs

Short version: the dock auto-empty fan is currently specified in volts, watts and kPa, and there is no airflow figure anywhere in the dock spec. contributions/vacuum-fan/README.md already argues that Pa is the wrong metric — so I tried to derive the right one for the dock.

Using only decisions the project has already frozen — the ID 29 mm tube in oomwoo-one-cad/docs/SPEC.md, measured from the X20 teardown — plus dilute-phase conveying velocity for coarse household debris:

≈ 13.2 l/s — 47.6 m³/h — 28 CFM

Checked against that, the 21.6–25.2 V / 65 mm / 350 W stick-vac class already named in BOM.md delivers ≈ 80 m³/h free-flow — roughly 1.7× the requirement, corroborated by two unrelated manufacturers. So the "weak-ish compared to consumer auto-empty docks" note in the BOM may be too pessimistic.

Two things I'd flag rather than bury:

• Those are free-flow numbers. The operating point is on the PQ curve, and the system curve can't be drawn yet — SPEC.md routes makeup air through the robot's own intake, so resistance is intake → bin → trap door → tube → cyclone in series. That puts fan selection downstream of the dust-bin geometry (RFC (Request for Contribution) - design, 3D print vacuum's dust bin component #2), not beside it. @oldocs, this may end up mattering for the bin's bottom port.
• Not every motor in the 65 mm stick-vac band passes — two variants land below or exactly at the requirement. The 350 W qualifier in the BOM is doing real work.

Also included: a Nidec vacuum blower line-up (51 pp.) that isn't in the Already found list, low-voltage candidates with published airflow, and two part numbers from BOM.md (Nidec 13F704P640, 64XC216-085D) that I could not find in any Nidec catalogue or anywhere online — would be glad to know where they came from.

Caveat, stated in the document too: nothing here was measured. Everything is vendor-published with provenance per datum, so it doesn't meet the "verifiable by someone else with the same part" bar on its own. It's meant as a baseline that tells whoever buys a part what to measure and what to compare it against. Happy to have any of it cut or corrected.