Worm Gear Reducer

Mount a single-start worm meshing with a worm wheel. Count the wheel's teeth — say 40.

Mark both. Turn the worm exactly one revolution and check the wheel: it should have advanced exactly one tooth.

Ratio = wheel teeth / worm starts = 40/1 = 40:1. No arithmetic about pitch diameters — the ratio is a tooth count and a start count, nothing else.

Now turn the wheel and try to make the worm rotate.

With a shallow-lead single-start worm, expect it not to move at all, however hard you push.

That is self-locking, and it is a friction effect, not a mechanical interlock — there is no catch or pawl anywhere in the assembly. It emerges from the thread angle alone.

Measure the worm's lead (axial advance per revolution) and its pitch diameter.

tan λ = lead / (π × pitch diameter)

Self-locking requires roughly tan λ < μ, the coefficient of friction — in practice a lead angle under about 5°.

Compute yours. If it is well under 5° it will self-lock; nearer 10° and it will back-drive. This is the whole design decision in one number.

Run the drive under load for two minutes, then feel the housing near the mesh.

Expect it to be noticeably warm — much warmer than a spur gear pair doing the same work.

The worm thread slides along the wheel tooth rather than rolling across it, and sliding friction is the loss. Every degree of that heat is efficiency you paid for the self-locking.

Note the materials: the worm is normally hardened steel, the wheel bronze.

That pairing is deliberate. Two hard surfaces sliding against each other under load gall and seize; a softer bronze wheel beds in to the steel worm, carries an oil film, and wears preferentially.

The wheel is designed to be the sacrificial part — cheaper to replace than the worm, and it protects the more expensive component. Same reasoning as the brass synchro ring in a gearbox.

Very old, and no single inventor. The worm and wheel is ancient — Archimedes is associated with the screw, and worm gearing appears in Hellenistic and Roman devices, in cotton gins, and throughout medieval and Renaissance machinery. Unlike the rest of this batch there is no foundational patent to cite, because the mechanism long predates the patent system. This blueprint therefore documents the principle rather than a claim, and does not manufacture an inventor for it.

Why it is in a power-transmission set anyway. Because it is the one common drive whose defining feature is a refusal to transmit — the only one specified for what it will not do.

Where the self-locking is the entire product. Hoists and winches that must hold a suspended load with no power and no brake. Machine-tool dividing heads. Gate and lift-bridge drives. The old worm-and-sector steering box, where self-locking meant road shocks did not come back through the wheel — and which was eventually abandoned precisely because drivers wanted some feedback.

The honest caution. Self-locking is a friction effect, so it degrades. Vibration can walk a self-locking worm drive backwards over time, because vibration momentarily reduces the effective friction. Anything holding a genuinely dangerous load — a passenger lift, a stage rig — carries a separate mechanical brake as well. Never treat self-locking as a safety device on its own.