Broaching Machine

One number governs the whole tool: how much taller each tooth is than the last.

1. Decide the total depth to be cut — say 3 mm for a keyway.
2. Choose a rise per tooth of 0.05 mm for soft material.
3. Divide: 3 ÷ 0.05 = 60 cutting teeth.
4. Add several finishing teeth at the FULL height with no rise, to size and burnish.
5. At 6 mm pitch, work out the total length — around 400 mm.

Rise per tooth is the feed, built into the tool. On every other machine in this batch the operator or a leadscrew sets how deep each cut goes. Here it was set when the broach was ground, and it cannot be changed. That is why a broach is expensive and why it is completely reliable: there is no setting to get wrong.

Sixty teeth by hand is unreasonable. Ten teeth proves the principle.

1. Take a 250 mm length of 12 mm aluminium flat bar.
2. Mark ten tooth positions at 12 mm pitch.
3. File each tooth to a progressive height, rising 0.1 mm per tooth.
4. Give each tooth a rake face and a clearance angle behind.
5. Measure every tooth height with the caliper and record the actual progression.

Measure and record the real heights, not the intended ones. A hand-filed broach will have an irregular progression, and knowing it lets you interpret the cut afterwards — a tooth that is too tall takes a heavy cut and may leave a mark you can trace back to it.

The broach is pulled, not pushed — and the work must be supported around it.

1. Build a frame with a fixed plate carrying a bushing that guides the broach.
2. Bore the bushing to fit the workpiece's existing hole, so the broach enters aligned.
3. Attach a pulling head to the broach's front end via a slot and pin.
4. Pull with a lever or a screw, using M8 × 100 hex bolts × 2, M8 flat washers × 4 and M8 hex nuts × 2 in the frame.

Pulling rather than pushing is a buckling decision. A long slender bar in compression buckles; in tension it cannot. That single fact decides the architecture of every broaching machine, and it is why short broaches may be pushed but long ones never are — the same reasoning that puts a chain in tension rather than a rod in compression.

The comparison that explains why the process exists.

1. Broach a keyway in a soft test piece in one pull. Time it.
2. Now cut an identical keyway on the shaping machine, stroke by stroke. Time that too.
3. Measure both for width, depth and parallelism.
4. Repeat both five times and compare the spread of the measurements.

Broaching became a production process in the early twentieth century, driven by the motor industry's need for splines, keyways and non-round holes in very large numbers. Early broaches were pulled by hand or by screw; hydraulic broaching machines followed and made the process fast and controllable.

It is the far end of a trend running through this whole batch. The planer put the tool's PATH into the machine. The turret lathe put the tool SEQUENCE into the setup. The hobbing machine put the tooth FORM into a rolling relationship. Broaching goes furthest: the entire process — roughing, cutting, finishing, sizing — is ground into one tool, and the machine does nothing but pull it in a straight line. Skill and judgement have been moved, step by step across a century, out of the operator and into the tooling.

What it can do that nothing else can: internal square, hexagonal, splined and irregular holes, in one pass, to a consistent size. A shaper can cut them slowly, a mill often cannot reach inside at all, and a grinder cannot get a wheel in there. For internal forms in quantity, broaching has essentially no competitor.

Its honest limits: one broach, one shape, one size — change the dimension and you buy another. High tooling cost, so unusable at low volume. It needs a starting hole, and it cannot cut a blind feature because the broach must pass right through. And a broken broach is expensive, usually caused by packed chips rather than by cutting force.