The CNC Machining of Mold, Explained

• 1
  What a mold actually asks of the cutting tool
• 2
  Why five axes change the mold cutting strategy
• 3
  Roughing, semi-finish and the tolerance budget
• 4
  How mold material changes feeds and tool life
• 5
  EDM and the features CNC cannot reach
• 6
  What the cutting accuracy means for mold life
• 7
  Matching the cutting method to the mold feature

A mold is not one shape. It is a cavity, a core, a shut-off face, a gate, a cooling circuit and a stack of plates that must close to within a few hundredths of a millimeter, cycle after cycle. The cnc machining of mold is therefore not a single operation. It is a sequence of operations, each with its own tool, its own tolerance budget and its own failure mode.

The part geometry decides almost everything. A shallow open pocket with generous radii can be cut in one setup on a 3-axis mill. A deep ribbed core with undercut side actions cannot. The tool has to reach the floor of the cavity without the holder crashing into the wall, and that reach requirement is what pushes a job from 3-axis to 5-axis.

On a 3-axis machine the tool axis is fixed vertical. The only way to cut a steep wall is with a small stepover, which means many passes and a long cycle. On a simultaneous 5-axis center the tool tilts so the flank of the cutter engages the wall at a consistent angle. Stepover can grow, and the cycle gets shorter.

Tilting also solves reach. A long slender tool deflects under cutting load, and deflection shows up as chatter marks on the mold surface. By tilting the tool and using a shorter gauge length, the same feature can be cut with a stiffer setup. That is usually the difference between a surface that polishes in two hours and one that fights back for two days.

There is a cost side too. Five-axis toolpaths are harder to verify, post-processors are more sensitive, and a crash is more expensive. For simple prismatic pockets and flat shut-offs, 3-axis remains the faster and cheaper route. We run 16 simultaneous 5-axis centers, 12 four-axis mills and 27 three-axis machines, and jobs are routed to whichever class actually needs the capability.

Roughing removes the bulk of the stock and leaves a uniform allowance. On a typical P20 or 1.2344 block, a face mill or a 16 mm indexable end mill takes the bulk down to roughly 0.5–1.0 mm of stock on the cavity walls. Leaving too little stock risks a hard skin from the previous cut; leaving too much loads the semi-finish tool and shortens its life.

Semi-finish brings the allowance to 0.1–0.3 mm and, more importantly, makes it even. A finishing tool follows the surface it is given, so any uneven allowance shows up as a visible witness line after finishing. This is the step where most shops cut corners, and it is the step that decides whether the mold polishes cleanly.

Finishing cuts the final surface. For a general mold surface we work to ±0.005 mm (±0.0002 in) where the drawing calls for it, with Ra 0.8–1.6 μm as a typical as-machined finish. Where a finer surface is needed, Ra 0.2–0.8 μm is reachable, but it is usually faster to machine to Ra 0.8–1.6 μm and polish the last bit by hand than to chase the finish with the cutter.

Thermal stability matters more than most people expect. A block that heats up during a long roughing cycle moves. If the finishing cut starts while the block is still warm, the dimensions shift as it cools. We inspect 100% of parts before shipment, but inspection cannot fix a dimension that moved after the cutter left.

Pre-hardened P20 and 1.2738 cut freely at 28–34 HRC and are the normal choice for prototype and low-volume tools. Hardened tool steels such as H13 (1.2344) at 48–52 HRC cut much slower, need coated carbide and a rigid setup, and punish any tool that is not running true. S136 and S136H are chosen when corrosion resistance matters, for example in medical or food-contact molds.

Aluminum molds, usually 6061 or 7075, machine fast and are common for prototype runs and bridge tooling. They wear quickly and do not hold a sharp shut-off edge for long, but they can be cut in a fraction of the time. Copper alloys are used for inserts where higher thermal conductivity helps the cooling cycle.

The practical limit is not the material itself but the aspect ratio of the feature. A 3 mm cutter in a 30 mm deep rib is a 10:1 reach. In hard steel that tool will deflect. The answer is a 5-axis tilt, a shorter tool, or an EDM electrode, and the choice usually comes down to how many cavities the tool has to survive.

Every cutter has a radius. A sharp internal corner, a narrow slot floor or a fine rib detail smaller than the smallest available cutter has to be cut by EDM instead. That means a machined electrode, usually copper or graphite, cut to the same tolerance class as the mold itself.

This is why the two processes sit side by side rather than competing. CNC removes the bulk quickly and holds the large surfaces. EDM takes the corners, the deep narrow details and the sharp edges. A mold shop that can only do one of them will either burn cycle time on the wrong process or lose detail.

The handoff point is roughly when the feature width drops below the smallest rigid cutter for the depth involved. As a rule of thumb, once the depth-to-width ratio passes about 5:1 in hardened steel, we plan the electrode into the process from the start rather than discovering the problem at the machine.

Mold life is decided by a handful of dimensions, not by the whole surface. Shut-off faces control flash. Gate detail controls fill and cosmetic quality. Cooling channel position controls cycle time and warpage. These are the features where ±0.005 mm matters and where a sloppy fit shows up as a production problem weeks later.

A well-cut cavity also reduces manual work. Consistent scallop height means the polisher removes a known amount of material, not an unknown one. That protects the geometry the cutter worked to hold. In practice, a mold that machines clean often ships days earlier because the bench work shrinks.

For the buyer, the useful question is not how many axes a shop owns. It is whether the shop can hold the tolerances on the features that decide mold life, and whether it knows when to hand a feature to EDM instead of forcing a cutter into it.