Gear Hobbing

Note the core: tool and workpiece run coupled, and the tooth form appears as an envelope. No dividing head, no tooth shape in the cutter.

Cut a straight rack from cardboard: 20 mm pitch, 20° flank angle, 12 mm high. That is your tool.

Cut a 127 mm diameter disc from cardboard. Circumference = 20 mm × 20 teeth.

Pin the disc to the board through its centre with a machine screw so it rotates freely.

Draw a line tangent to the disc. The rack must travel exactly one pitch while the disc turns by one tooth.

Advance the rack 5 mm, turn the disc by the matching amount, and trace the tool outline onto the disc each time.

Repeat until the disc has gone once around. Between the lines the involute form appears — it was never drawn, it was generated.

Cut the disc along the inner envelope curve. You are holding a generated gear.

Measure tooth centre to tooth centre with the caliper at five places. Record the largest deviation.

Make a second gear of 10 teeth from the same rack. The two must mesh with no rework — that is the real proof.

Mark one tooth on each and turn. 20 teeth to 10 gives exactly two turns to one. Count them.

Draw a gear made with a dividing head: 20 separate indexings. Every indexing error stays in the gear. Generating has no indexing step to get wrong.

Examine an HSS end mill under the magnifier. Its edge has nothing to do with a tooth form — which is precisely the point.

US 640,728, “Apparatus for Cutting Worm-Wheels”, Robert Hermann Pfauter, Chemnitz. Filed 10 August 1899, granted 2 January 1900. The German application of 1897 is regarded as the original; its DRP number could not be verified and is therefore not asserted here. Where a number is missing, the function is described — not guessed.

What came before. Christian Schiele described the generating principle in 1856, and Ramsden was working along similar lines on dividing engines as early as 1768. Pfauter's contribution is the ready-to-use, universal machine from 1897 in Chemnitz; the Pfauter company led gear hobbing for over a century and is today part of Gleason.

Why the process won. A single hob of a given module cuts any number of teeth. With indexing you need the matching form cutter for each tooth count, and every indexing step contributes its own error. Generating is faster, more accurate and cheaper all at once — which is rare.

The link to mass production. Without accurate, cheap gears there is no gearbox, and without gearboxes no automobile in quantity. Hobbing stands at the head of that chain. The Fellows gear shaper (1897, USA) solves the same problem by shaping rather than milling — both processes are still in use side by side today.