Guide to CNC Machining Magnetic Shields

• 1
  Key takeaways
• 2
  How a magnetic shield actually works
• 3
  Alloys used in CNC machining magnetic shields
• 4
  Geometry, seams, and why microns matter
• 5
  Machining high-permeability alloys without wrecking them
• 6
  When CNC machining is the wrong process
• 7
  A workable sequence for a machined shield
• 8
  Material and process choices at a glance

A magnetic shield does not block a field the way a wall blocks light. It gives the flux an easier road. High-permeability alloys offer far lower reluctance than air, so field lines bend into the shield wall, travel along it, and rejoin outside. The protected volume in the middle sees a weaker field than it would in free space.

Two mechanisms matter, and they peak in different frequency bands. At DC and low frequency, the dominant effect is flux diversion through the wall material. Above roughly 10 kHz to 100 kHz, depending on wall thickness and conductivity, eddy currents in the same wall start to oppose the field, and the shield begins to behave like a conductive screen.

That split explains why a single material rarely solves everything. A 1 mm Mu-metal wall may give excellent DC attenuation but poor high-frequency performance if it is also a poor conductor. A copper layer adds eddy-current shielding but does nothing for a static field. Many real enclosures stack both.

The number you care about is attenuation, usually quoted in decibels or as a ratio of external to internal field. It is not a fixed property of the alloy. It depends on wall thickness, the largest opening, the distance from the source, and how well the joints conduct flux.

• 1
  Low frequencyMaterial permeability and wall thickness dominate.
• 2
  High frequencyConductivity and skin depth dominate.
• 3
  Both bandsOpenings and seams set the practical ceiling.

Mu-metal is the common name for a nickel-iron alloy near 80% Ni with iron and a few percent of molybdenum or copper. Its initial permeability can reach very high values, but only after a hydrogen anneal. That anneal is what unlocks the performance, and it is also why machining sequence matters so much.

Permalloy covers a family of Ni-Fe alloys, typically 45% to 80% Ni. The 80/20 grades behave close to Mu-metal. Lower-nickel grades are cheaper, easier to form, and more tolerant of mechanical stress, at the cost of lower peak permeability. For a shield that only needs moderate attenuation, a 50% Ni grade is often the sensible choice.

We also machine plain 304 and 316 stainless and low-carbon steel for shields that sit in front of strong permanent magnets and only need gross redirection. These are not high-permeability materials. They are stiff, weldable, and cheap, and they work when the field is large and the attenuation target is modest.

The list below is what we run in our shop, not a claim that each grade suits every shield.

One warning applies to all of them. Cold work from milling, turning, or bending raises dislocation density and drops permeability. A heavy roughing pass can cut effective permeability by a large factor. Annealing after machining is not optional for high-permeability parts.

• 1
  Nickel-ironMu-metal and 80/20 Permalloy for the highest attenuation.
• 2
  50% Ni PermalloyLower cost, easier to machine, moderate attenuation.
• 3
  Stainless and carbon steelMechanical protection, low permeability, weldable.
• 4
  Copper and aluminiumEddy-current layer for the high-frequency band.

Flux prefers a smooth, continuous path. A scratch, a step, or a raised burr on the inner surface raises local reluctance, and field lines leave the wall early. That is the practical reason surface finish on the inside of a high-permeability shield matters more than on the outside.

Gaps are worse than roughness. A 0.1 mm air gap at a flange corner can dominate the total reluctance of a joint. Where a shield must be split for assembly, the joint should be lapped or ground flat, then closed with enough fasteners to keep the mating faces in contact across the full seam length.

Wall thickness sets two things at once. More material means more flux-carrying cross section, so attenuation improves. It also adds stiffness, which helps thin shells survive machining. The trade-off is weight: aerospace and satellite assemblies often push toward 0.5 mm or thinner, and those walls need fixturing rather than heavy cuts.

Apertures are the third limit. Every cable feedthrough, vent, or optical window is a hole in the flux path. Keep openings small, orient them away from the sensitive volume, and where possible add a short tube or a second shield layer around the opening.

• 1
  RoughnessKeep inner surfaces at Ra 0.8–1.6 μm or finer.
• 2
  SeamsLap or grind mating faces; avoid gaps over 0.05 mm.
• 3
  AperturesSmall, few, and pointed away from the payload.

Nickel-iron alloys work-harden quickly and conduct heat poorly. A sharp, positive-rake carbide tool with a generous edge radius cuts cleanly. A dull tool rubs, and rubbing is the enemy. Change inserts on a fixed count rather than waiting for a poor finish.

Light radial passes beat heavy ones. On a 1 mm wall, a 0.2–0.3 mm radial stepdown with a 6 mm end mill keeps cutting forces low and heat moving out with the chip. Flood coolant, not mist, because the material holds heat near the cut. Climb milling on the finishing pass gives a better inner surface.

Thin walls deflect before they chatter. Support them with a soft-jaw fixture, a wax or low-melt fixturing compound, or a sacrificial boss that gets removed last. For a 0.5 mm wall over a 100 mm span, we plan the operation sequence around the fixture, not around the toolpath.

Five-axis machining earns its place here. A single setup lets the tool reach the inside of a deep pocket, the flange face, and the mounting bosses without re-clamping. Each re-clamp is a chance to bend a thin part and a chance to lose position. Fewer setups means fewer opportunities for damage.

• 1
  Roughing0.5–1.0 mm radial stepdown, flood coolant, sharp inserts.
• 2
  Finishing0.1–0.2 mm radial pass, climb milling, Ra 0.8–1.6 μm.
• 3
  Tolerance±0.005 mm on critical mating faces and bore locations.
• 4
  After machiningHydrogen anneal to restore permeability, then clean.

Machining is not always the answer. If the part is a thin-walled cylindrical can with a closed end and a length more than about three times its diameter, deep drawing is cheaper and avoids the seams that kill attenuation. Machining that shape means boring a deep pocket and then fighting distortion.

Large curved shells follow the same logic. A spherical or parabolic shield over 300 mm across is usually formed and welded, then annealed. Machining it from solid wastes material and leaves internal tool marks that need polishing.

Very high volume also pushes away from machining. Above a few thousand identical enclosures per year, a progressive die or a die-cast housing with a plated finish will beat a milled part on unit cost. Machining wins on prototypes, low volume, tight tolerances, and geometry that a die cannot release.

There is also a hard physical limit. Machining cannot anneal. If your supplier cuts the part and ships it without a controlled anneal, the measured permeability may be a fraction of the datasheet value, and the shield will not meet its attenuation target no matter how good the tolerances are.

• 1
  Deep thin cansUse deep drawing, not boring.
• 2
  Large curved shellsForm and weld, then anneal.
• 3
  High volumeDie casting or stamping above a few thousand units.