CNC Machining Magnets: Where They Help and Where They Fail

• 1
  CNC machining magnets act in two directions
• 2
  How magnetic workholding actually holds a part
• 3
  Magnetic chucks, sine plates and demagnetizing
• 4
  When a magnet is the part being machined
• 5
  Where magnet workholding is the wrong answer
• 6
  What to put on the drawing and in the inspection plan
• 7
  Magnet workholding versus other methods

There are two separate questions behind the phrase CNC machining magnets. The first is whether a magnet can hold a part while it is cut. The second is whether the shop can cut a magnet itself, into a small, accurate, finished shape. They share a word and almost nothing else.

Direction one is workholding. A permanent magnet or an electromagnet pulls a ferromagnetic workpiece down onto a table, with no clamps in the cutting zone. That matters on thin plates and on parts where a clamp would sit exactly where the tool needs to go. Direction two is part production. NdFeB, SmCo and ferrite blanks are brittle and heat-sensitive, and turning them into a finished rotor or pole piece is a grinding and wire-EDM problem more than a milling problem.

Keep the split in mind for the rest of this page. Workholding questions are about clamping force, wall thickness and residual magnetism. Part-making questions are about material grade, edge chipping, coolant and how the geometry is inspected. Mixing the two leads to fixtures that slip and magnets that crack.

A magnet only develops force against a material it can magnetize. Low-carbon steel, 1018 and 1045, most ferritic stainless, and cast iron respond well. Austenitic stainless such as 304 and 316, aluminium, copper, brass, titanium and every plastic in our material list respond essentially not at all. Hold a 304 plate on a magnetic chuck and it will slide under the first heavy cut.

Force also drops with thickness. A 1 mm steel shim saturates quickly, so the flux path closes through the plate and gives limited holding force. A 10 mm plate carries far more. As a rough working rule, magnetic workholding becomes practical on low-carbon steel above about 3–5 mm, and marginal below 2 mm unless you add a mechanical stop or a backing plate.

The third variable is the air gap. Scale, a burr, a sheet of paper or a coat of paint between the pole and the part can cut holding force by half. Clean the pole faces and stone the part's bottom face before you trust the setup. With a Ø400 mm rotary table and a magnetic chuck on a 4-axis machine, that cleaning step is what separates a stable run from a scrapped plate.

Surface finish matters too. A ground or milled flat face gives a good flux path. A rough sawn or heavily pitted face does not. If the stock arrives with mill scale, face it first, even if the drawing does not require a finished face. You are buying holding force with that cut.

Magnetic chucks come in two families. Permanent magnetic chucks use a lever or key to switch the flux path on and off, and they need no power to stay engaged. Electro-permanent chucks use a short electrical pulse to change state, then hold with no current. Both are common on surface grinders and on 3-axis mills running flat plates. Neither works through a non-magnetic sub-plate unless the chuck is designed for it.

For grinding, a fine pole pitch chuck is usually the right call. Closer poles give a denser field and hold thin parts flatter, which is what you want when the tolerance is ±0.005 mm. A coarse pole chuck on a 1 mm washer will bend the part into the field and spring it back after release. The flatness you measure on the chuck is not the flatness you ship.

After cutting, the part often keeps a small field. That residual magnetism attracts chips, upsets touch probes and can confuse a Hall-effect sensor during the next assembly step. A demagnetizer cycle takes seconds and belongs in the process sheet for any steel part that was held magnetically. Soft magnetic steel such as 1018 demagnetizes easily. Hardened 440C and some tool steels hold more and may need two passes.

Do not assume a magnetic chuck replaces a vise for every flat part. On a 5-axis job with heavy side milling, the cutting force has a large tangential component that a magnet resists poorly. A magnet is at its best when the dominant load is downward and the part is thin.

Sintered NdFeB blanks are hard, brittle and electrically conductive. Milling them with a standard end mill tends to chip the edges and can heat the blank past its working temperature, which permanently reduces its magnetic output. The usual route is diamond grinding for flat faces and wire EDM for profiles and slots. Both are cold-cutting in the sense that matters: little heat reaches the magnet body.

SmCo blanks tolerate higher temperatures and are often chosen for aerospace and high-temperature sensor work. They are also more brittle than NdFeB and more expensive. Ferrite is cheap and stable but soft in magnetic terms, so it needs a larger volume for the same flux. Which one you pick should come from the operating temperature and the flux requirement, not from the drawing alone.

The magnet grade, for example N42 or N52, is part of the specification and should be printed on the drawing. If only a shape is supplied, we cannot verify output. Ask the designer for grade, coating and the magnetizing direction before quoting, because a magnetized blank is harder to handle and cannot be shipped by every route.

Coatings matter for corrosion and for handling. Nickel-copper-nickel is the common default. Epoxy and zinc are used where the part sees humidity or salt spray. After machining, an uncoated NdFeB edge corrodes quickly, so plan the coating step into the sequence rather than adding it later.

If the part is austenitic stainless, aluminium, titanium, Inconel or plastic, a magnetic chuck does nothing. That covers a large share of the work we see for aerospace and medical devices. For those parts, use a vacuum chuck, a fixture plate with toe clamps, or glue-fixturing on a sacrificial plate and machine the setup off in a second operation.

If the part is magnetic but thin and unsupported, the magnet can pull it out of flat. Thin steel washers, shim stock and laminated stacks are the classic cases. A backing plate of the same steel grade, ground flat, spreads the load and keeps the part from being drawn into the pole gaps.

If the operation is heavy roughing with a large radial depth of cut, magnetic holding is usually not the right primary clamp. Use it as a support under a vise or as a secondary locator, and let a mechanical clamp take the side load. This is a judgement call, not a rule, but the failures we see come from ignoring it.

If the finished part goes into a sensor or a magnetic assembly, residual magnetism is a real risk. Add a demagnetizing step and specify a maximum residual flux in the inspection plan. Do that at quoting time, not after the first batch.

For a machined magnet, the drawing should carry the grade, the coating, the magnetizing direction and the tolerance that matters. Dimensions on a brittle part are best held by grinding or wire EDM, so a tolerance tighter than ±0.05 mm on an NdFeB profile adds cost fast. Say which faces are functional and which are clearance.

For a magnetically held part, the process sheet should say how it was clamped, whether it was demagnetized, and what residual flux is acceptable. Without that note, the next operation inherits a problem nobody can trace. We inspect 100% of parts before shipment and can supply reports on request.

Tolerance capability on our machines is ±0.005 mm, with finishes from Ra 0.2–0.8 μm on ground faces to Ra 1.6–3.2 μm as machined. Those numbers describe what the machine can hold, not what every magnet job needs. Over-specifying finish on a non-functional face is the most common way to pay for nothing.

We run ISO 9001:2015, IATF 16949:2016, ISO 13485:2016 and ISO 27001:2022. For sensor and motor work, the IATF and ISO 13485 scopes are usually the ones a customer's quality team asks about.