CNC processing: the revolution in mold manufacturing

• 1
  What mold making looked like before CNC processing
• 2
  How CNC processing the revolution changed the physics of cutting
• 3
  Where five-axis motion earns its cost in mold work
• 4
  What CNC still cannot do in mold manufacturing
• 5
  How a CNC mold job runs on the floor
• 6
  Choosing a cutting route for a mold cavity

Manual mold making was a chain of handoffs. A pattern or a graphite electrode was cut on a tracer mill that followed a template. Then a die sinker burned the cavity. A fitter stoned the parting line and spotted the shutoffs by hand. Each step added its own error.

The template itself wore down. Graphite electrodes eroded unevenly during the burn, so the cavity drifted from the drawing a few hundredths of a millimeter at a time. Nobody knew the final dimension until the mold was assembled and a first-shot sample was measured.

That measurement usually came back wrong. Corrections meant welding, remachining or cutting a new electrode, and the loop started again. Two to four correction loops were normal on a complex part.

This is why tooling schedules of six to twelve weeks were accepted as normal. The time was not spent cutting metal. It was spent finding out where the metal ended up.

CNC removed the template from the loop. The part geometry lives in a tool path file, and the machine follows that file with closed-loop feedback on each axis. The cutting edge no longer depends on a physical master that can wear.

The second change was rigidity. A modern machining center holds the tool in a spindle that resists deflection under load. When the tool does not bend, the programmed depth of cut is the real depth of cut. That is what lets a shop hold ±0.005 mm across a cavity instead of chasing it after assembly.

The third change was the tool path itself. CAM software can drive a constant chip load, ramp into the material instead of plunging, and keep the cutter engaged at a stable angle. Tool life goes up, and so does surface quality.

Together these three shifts moved the correction loop out of the mold shop and into the software, where a change costs minutes instead of days.

• 1
  Geometry sourceDigital tool path replaces a wearing template or electrode.
• 2
  StiffnessLow deflection keeps the programmed dimension as the real dimension.
• 3
  Chip load controlStable engagement improves both finish and cutter life.

A three-axis machine reaches a cavity from one direction. Every undercut, every deep rib and every steep wall has to be reached by a long tool or by a second setup on a different face. Long tools chatter. Second setups lose alignment.

Five-axis machining tilts the tool or the table so the cutter approaches the surface at a controlled angle. Short, stiff tools can reach deep pockets. Steep walls get cut with the flank of the tool instead of the tip, which spreads the wear and leaves a cleaner surface.

The practical gain is in two places. First, deep ribs and narrow slots that would need EDM on a three-axis route can be milled directly. Second, the number of setups drops, so the accumulated position error between operations drops with it.

Five-axis is not free. Programming takes longer, and a poorly posted tool path can scrape a cavity faster than a three-axis cut ever would. It pays off on complex cores, tall ribs and parts with many angled faces. On a flat plate with simple pockets, a three-axis machine is faster and cheaper.

CNC is a subtractive process with a spinning cutter. That sets two hard limits. The first is corner radius: an internal corner can never be sharper than the tool that cut it. A Ø6 mm end mill leaves a 3 mm corner radius. If the drawing calls for a sharp internal corner, someone still has to burn it or file it.

The second limit is hardness. Cutting tools will machine pre-hardened steel in the 30–40 HRC range with the right grades and light depths of cut. Past that, the economics flip and the cavity is better roughed soft and finished by EDM or by hard milling on a dedicated machine.

Surface finish has a limit too. As-machined surfaces land around Ra 1.6–3.2 μm. High-finish passes reach Ra 0.8–1.6 μm, and fine finishing can reach Ra 0.2–0.8 μm. Below that, polishing or lapping takes over.

Aspect ratio matters as well. A slot that is 10 mm wide and 100 mm deep will deflect any tool that fits inside it. In those cases, EDM or a different design is the honest answer, not a longer end mill.

A job starts with DFM. Our engineers review the mold design for tool reach, corner radii, draft and accessible faces, then send back notes with the quotation. This happens within 12 hours, and it is where most cost is either saved or locked in.

Roughing removes bulk material with a large cutter and heavy depths of cut, leaving a uniform stock allowance for the finishing passes. On a hardened core, this is done before heat treatment so the finishing cut only removes a light skin.

Semi-finishing and finishing follow, often on a five-axis center so the tool can tilt into steep walls. A Ø400 mm rotary table handles round cores and slides that would otherwise need multiple setups.

Final inspection checks the cavity against the model. We inspect 100% of parts before shipment, with raw material checks, in-process monitoring and a final report on request. For a mold, that report is what tells you the cavity is ready for a first shot.