Treatment of Mold Electric Sparks: How EDM Actually Removes Metal

• 1
  What Happens in the Gap During Treatment of Mold Electric Sparks
• 2
  How Pulse Settings Control the Cut
• 3
  Powder-Mixed Dielectric and Fine Finishing
• 4
  Where EDM Fits in a Mold Shop
• 5
  Recast Layer, Electrode Wear, and Inspection
• 6
  EDM vs High-Speed Milling: What Changed
• 7
  Mold Insert: EDM or Milling First

Electric discharge machining removes metal with heat, not with a cutting edge. The electrode and the workpiece never touch. A servo axis holds a gap of roughly 0.01–0.05 mm, the dielectric fluid breaks down, and a plasma channel forms for a few microseconds. The channel reaches several thousand degrees Celsius, melts a small crater on the workpiece, and the fluid flushes the debris away before the next pulse.

That single mechanism explains most of the process behavior. Because removal is thermal, hardness does not matter much. A 52 HRC mold insert cuts about as easily as the same steel in the annealed state. This is why the treatment of mold electric sparks is routine after heat treatment, when milling and drilling become slow or impossible.

It also explains the limits. The recast layer on the cut surface is not base metal. It is rapidly solidified material with different hardness and, in some steels, micro-cracks. On a mold cavity that sees high injection pressure or a sharp corner, that layer matters. For most cores and slides, a light finishing pass removes enough of it to be safe.

The gap has to stay clean. If flushing is weak, debris collects, pulses become unstable, and you get arcs instead of controlled discharges. Arcs cut fast and rough, and they leave a burned pit that is hard to polish out. On deep ribs and narrow slots, flushing is the real constraint, not the generator.

Two numbers set the outcome: peak current and pulse on-time. Raise the current and the craters get bigger, so material removal rate climbs and surface finish gets coarser. Shorten the on-time and each discharge carries less energy, so the surface gets finer but the cut slows down. Roughing and finishing are the same process at different points on that curve.

A common sequence for a hardened mold insert runs three passes. Roughing uses high current and long on-time to move volume, leaving Ra 3.2–6.3 μm. Semi-finish drops the current and shortens the pulse. Finish passes use low current, very short on-time, and a small gap, reaching Ra 0.2–0.8 μm. Each pass also uses a smaller electrode offset, because the gap narrows as energy drops.

On-time has a ceiling. Past a certain duration the discharge stops behaving like a spark and starts behaving like an arc. The plasma channel stops collapsing between pulses and stays in one spot. The result is a local pit, carbon build-up on the electrode, and a surface that will not clean up. Modern generators cap the on-time automatically, which is one reason older machines are harder to run on fine work.

Duty cycle links the two. It is on-time divided by the total cycle. Low duty cycle gives the gap time to deionize and flush. High duty cycle raises removal rate but raises the risk of unstable discharge. On deep cavities, keep duty cycle low and accept a slower cut.

Standard dielectric is hydrocarbon oil or deionized water. It insulates, cools, and carries debris out of the gap. Add fine powder to the oil and the behavior changes. Silicon or aluminum powder suspended in the dielectric lowers the breakdown resistance, so the discharge spreads into many small, shallow craters instead of a few deep ones.

The practical effect is a better surface at the same removal rate, or the same surface with fewer passes. Powder mixing is used mostly on mirror-finish mold cavities and on small, deep details where a final hand polish is difficult. It also reduces the risk of a concentrated arc, because the discharge energy is distributed.

The trade-offs are real. Powder-mixed fluid needs its own tank, agitation, and concentration control. If the powder settles or the concentration drifts, results drift with it. It also costs more to maintain than plain oil. For an ordinary core insert with Ra 1.6 μm requirement, the extra setup rarely pays back.

Electrode material interacts with all of this. Copper and graphite wear differently. Copper holds fine detail and gives a cleaner finish. Graphite cuts faster and is cheaper for large roughing electrodes, but it wears at corners. On tight ribs, many shops run graphite for roughing and copper for finishing, then accept the electrode change cost.

The process earns its place in three situations. First, hardened material. Once a mold insert is at 48–52 HRC, milling the cavity is slow and tool wear is heavy. EDM does not care. Second, geometry that a cutter cannot reach: sharp internal corners, deep narrow ribs, square-bottom pockets, and slots narrower than any available end mill. Third, surface detail that has to be burned in, such as texture or fine engraving.

It loses to milling everywhere else. A 5-axis machining center with the right toolpath will finish a soft mold plate faster and cheaper, and it will not leave a recast layer. For large, open cavities in pre-hardened steel at 30–38 HRC, high-speed milling is usually the better first choice. EDM is the second operation for the corners and details milling cannot finish.

Small holes are a separate case. A hole 0.3–3 mm across, drilled at an angle into a hardened slide, is a hole EDM job. A rotating electrode or a small wire guide does work a drill cannot reach without walking off center. On waterlines and ejector pin holes in hard steel, this often decides the process plan.

Part size matters too. Wire EDM handles a 4,000 mm maximum processing size on our large travel machines, but the taller the part, the more the wire has to flush and the slower the cut. Very tall, very precise work is where the process gets expensive.

Every EDM surface carries a recast layer, typically 2–20 μm depending on the finishing pass. The layer is harder and more brittle than the base metal, and in tool steels it can hold tensile stress. For a mold cavity under high injection pressure, that is a fatigue risk. Light finishing passes thin the layer, and a stress-relief temper after EDM removes most of the residual stress.

The rule we use is simple. If the surface will be polished, plan an extra 0.02–0.05 mm of stock so polishing removes the layer completely. If the surface is functional and stays as burned, keep the last pass light and specify a stress relief. Skipping both is how a mold insert cracks after a few thousand cycles.

Electrode wear is the other variable. Corner wear is worst because current density is highest there. On a finishing electrode, wear of 0.01 mm at a sharp corner shows up directly in the part. This is why finishing electrodes are often run at low current with an undersize allowance, and why some shops keep separate roughing and finishing electrodes instead of one stepped electrode.

Inspection closes the loop. We check the burned cavity against the electrode model, not just the drawing, because wear and gap offset both shift dimensions. On our EDM work the tolerance target is ±0.005 mm, with 100% inspection before shipment. Where a recast layer matters, we can supply surface reports on request.

High-speed milling took a lot of work away from EDM. With small-diameter tools, high spindle speeds, and CAM that keeps tool load constant, a 5-axis machine can finish a hardened cavity that used to require an electrode. On a medium mold insert in 40 HRC steel, milling is often faster and leaves a better surface for polishing.

What milling did not take is the geometry it cannot reach. Internal corners stay sharp only if the cutter radius allows it. Deep ribs need a long, thin tool that deflects. Texture and engraving still need a burn. And a repair on a hardened insert, where removing material is the goal, is usually an EDM job.

The result is that modern mold work is a split process, not a choice between two methods. Mill the open geometry, then burn the corners, ribs, and detail. On our floor, 16 simultaneous 5-axis machining centers handle the milling side, so an insert can be milled, heat treated, and then finished by EDM without waiting on an outside vendor.

When you are planning a mold, the useful question is not which process is better. It is which features need which process, and in what order. Getting that order right is what keeps a mold insert on schedule.