Why food separation is not an ordinary magnet job
In most applications a magnet that underperforms costs money. Here it can put metal into food. Three consequences follow from that, and they shape every specification decision on this page.
The failure mode that matters most
Thermal demagnetization from cleaning, not from the process. A separator specified against a 40 °C product temperature is frequently subjected to 80–90 °C clean-in-place cycles, steam sterilisation, or an oven pass that nobody included in the requirement. Standard N-grade NdFeB is rated to only 80 °C, and the loss it takes above that is irreversible — the field does not come back when the equipment cools.
The separator then runs indefinitely at reduced strength, still visually perfect, until an annual verification catches it or a metal fragment does. Specify the temperature class against the hottest thing that ever touches the magnet, including cleaning. The load-line method is in the temperature guide.
Separator formats by product form
Format follows the product stream. The wrong format cannot be rescued by a stronger magnet, because capture depends on the contaminant actually passing close to a magnet face.
| Format | Product form | How it captures | Watch for |
|---|---|---|---|
| Grate / grid | Free-flowing dry powders, granules, pellets | Product falls through a bed of magnetic tubes; contaminant is pulled to a tube surface | Bridging on cohesive powders; tube spacing versus particle size |
| Plate | Chutes, ducts, thin product streams | Suspended above or in the stream; captures on a flat face | Burden depth — anything above the working reach passes untouched |
| Liquid line / trap | Slurries, syrups, sauces, liquids | In-line tubes in the flow path | Pressure drop, dead zones, cleanability of the housing |
| Drum | High volume, continuous, heavy contamination | Rotating shell carries captured metal away from the product path | Self-cleaning is the point; verify the discharge actually clears |
| Pneumatic line | Product conveyed in air | In-line unit in the pipeline | Velocity — fast-moving particles may not deflect in the time available |
| Bullet / deflector | Gravity-fed streams in pipe | Cone in the flow path deflects product across a magnet face | Wear on the nose; ensuring full stream diversion |
Specifying the magnetic element
Field at the working surface, not grade
Separator performance is quoted in gauss at the tube or plate surface, because that is where capture happens. Grade alone tells you nothing — the same grade in a different pole geometry produces a very different surface field and a very different reach into the stream.
- Pole configuration sets the gradient. Capture force depends on the gradient of the field, not just its magnitude. Closely spaced alternating poles produce a steep gradient and strong capture close in; widely spaced poles reach further with less grip. Separator designers trade these deliberately.
- Rare earth versus ferrite is an application decision. Rare earth elements produce high surface fields and steep gradients suited to fine and weakly magnetic particles — work-hardened stainless fragments in particular. Ferrite reaches further into a deep burden with a gentler gradient and costs a fraction as much.
- Weakly magnetic contamination is the hard case. Austenitic stainless is nearly non-magnetic as supplied, but becomes measurably magnetic after cold work — a fragment from a worn screen or a broken blade. Capturing it needs high gradient and close approach, which is exactly what rare earth grate systems are for.
Temperature class
| Duty | Typical exposure | Minimum class |
|---|---|---|
| Ambient dry product, wipe-down cleaning | Up to ~40 °C | N — but leave margin |
| Warm product, hot water wash | 60–80 °C | M or H |
| Clean-in-place, caustic hot cycles | 80–100 °C | H |
| Steam sterilisation, hot oil, bake lines | 120–180 °C | SH or UH |
| Above that | >200 °C | Samarium cobalt territory |
Ask the cleaning question explicitly
Process engineers specify against product temperature; sanitation runs the equipment far hotter. The single most useful question at specification time is “what is the hottest cycle this equipment ever sees, and does the magnet stay in it?” Half the over-temperature failures in food separation are answered by that sentence.
Construction and finish
- The magnet is encapsulated, not exposed. Separator elements are sealed inside stainless tubes or housings — the food-contact surface is stainless, typically 316, with a specified surface finish. The magnet itself never touches product.
- Seal integrity is the failure point. A compromised weld or seal lets moisture reach the magnet stack, and NdFeB corrodes from the grain boundaries outward. Specify the seal and its verification, not just the tube material.
- Sanitary design is the housing designer’s domain — continuous welds, no crevices, drainable geometry, appropriate surface finish. Guidance frameworks such as EHEDG and 3-A cover this; it is separate from the magnet specification but constrains it dimensionally.
- Internal coating still matters. Even inside a sealed tube, specify the magnet coating for the environment it would see if the seal were breached. Detail is in certificates and compliance documents.
Validation, verification and the audit trail
This is where a food separator differs most from an industrial one, and where most audit findings land.
The major GFSI-recognised schemes — BRCGS, SQF, FSSC 22000 — all expect this cycle for physical hazard controls, built on Codex HACCP principles. For magnetic separators specifically, HACCP International publishes a dedicated standard (0909MAGSEP, currently in its 2021 revision) that certification bodies and specialist validators work to.
Gauss meter, not pull test
The current direction of practice is to verify separator strength with a calibrated gauss meter at defined measurement points rather than by pull testing. A pull test is a system measurement — it depends on the test piece, the contact condition, the operator and the pull direction, and it scatters far more than the underlying field. Two technicians pull-testing the same conforming separator can reach different conclusions.
A gauss reading at a defined point on a defined surface with a calibrated instrument is repeatable, comparable year over year, and defensible in an audit. If your programme still specifies pull testing, that is worth revisiting. The instrument-by-instrument reasoning is in how magnets are tested.
What a defensible magnet programme contains
- An inventory of every separator: location, format, installation date, and its role in the food safety plan.
- Baseline field measurements at defined points, taken at installation, so later readings have something to be compared against.
- Documented inspection and cleaning at a frequency justified by the risk, recording what was recovered — the trend in recovered metal is itself an early warning about upstream equipment.
- Periodic measured verification, commonly annual and commonly third party, with a calibrated instrument and a report retained.
- In-house interim checks between formal verifications, which is where a mid-year loss from an over-temperature event actually gets caught.
- A defined action limit — the field level below which the separator is removed from service — set from the validation, not chosen after a bad reading.
What we supply
Where we sit in this supply chain
Radial Magnets supplies magnet elements, stacks and assemblies to separation equipment manufacturers and to plants replacing elements in existing housings. We are not a certifying body for food safety systems, and a finished sanitary separator is the equipment builder’s assembly and responsibility. What we can do is make sure the magnetic element inside it is the right grade, the right pole geometry, and documented to the lot.
- High-coercivity NdFeB in H, SH and UH classes for washdown and elevated-temperature duty, where standard grades quietly demagnetize.
- Samarium cobalt for genuinely hot lines above the NdFeB range, and where corrosion resistance without coating is an advantage.
- Ceramic ferrite where deep-burden reach and cost matter more than gradient — still the right answer for many bulk streams.
- Multi-pole stacks and assemblies built to a specified surface field and pole pitch rather than to a grade alone.
- Per-lot material and coating certification, so the element inside a validated separator is traceable.
What to send us
- Product form and flow rate, and the burden depth at the separator position
- Contaminant type and size you need to capture — and whether work-hardened stainless is in scope
- Maximum temperature at the magnet, including cleaning and sterilisation cycles
- Cleaning chemistry and method
- The housing envelope, or the existing element dimensions if this is a replacement
- Required surface field and the measurement points, if your validation already defines them
- Certification the plant or the scheme requires