The Four Numbers on a Magnet Datasheet, and Which One Actually Matters

A permanent magnet datasheet has four numbers on it. Most buyers read the first one and the grade code, and neither is the number that predicts failure.

Here is what each one means and which one to argue about.

How strong the magnet is. Measured in tesla, typically 1.15 to 1.48 T for the neodymium grades used in traction motors.

This is the number that sets torque per amp and back-EMF per rpm. It is the one everybody quotes and the one that is easiest to understand. It is also not the number that will cause you a problem.

The reverse field needed to drive the magnet’s flux to zero. Frequently quoted, rarely useful. Skip to the next one.

This is the number that matters.

The difference between H_cB and H_cJ is the difference between cancelling a magnet’s flux and actually destroying its magnetisation. H_cJ is the field that genuinely damages the magnet, and it is always the larger of the two.

Why it matters: a magnet in a motor is under constant reverse field from its own geometry, plus whatever the stator throws at it. If that combined field pushes past a certain point — the knee of the demagnetisation curve — domains flip and do not flip back. The magnet is permanently weaker. Cooling it does not help. The only recovery is removing it and re-magnetising it in a saturating field, which for a bonded rotor means scrap.

The best single figure of merit, in kJ/m³, because it captures strength and resistance to demagnetisation together. It is what the grade number encodes.

Take N42SH:

• •N — neodymium.
• •42 — (BH)max in MGOe. Multiply by 7.96 to get kJ/m³, so 42 becomes about 334 kJ/m³.
• •SH — maximum operating temperature, 150 °C.

The letter suffixes run: no letter = 80 °C, M = 100, H = 120, SH = 150, UH = 180, EH = 200, AH = 230.

Two numbers change with temperature, and they change at very different speeds.

• •Remanence falls at about −0.12 per cent per °C.
• •Intrinsic coercivity falls at about −0.6 per cent per °C.

Coercivity degrades five times faster than strength. So a magnet does not gradually fade toward failure. It performs to specification, and to specification, and to specification — and then at some temperature it crosses the knee under a load it survived yesterday, and loses a permanent chunk in a fraction of a second.

And it needs two conditions simultaneously:

• •A hot magnet at a benign operating point is fine.
• •A cold magnet hit with a large reverse field is fine.
• •A hot magnet hit with a large reverse field is how rotors are destroyed.

The design case is specific and foreseeable: a three-phase short circuit at high speed with the rotor already at maximum temperature. An inverter fault at the end of a long gradient.

Two things, and a supplier who has neither has not done the work.

• •A hysteresigraph trace for your lot, at your maximum operating temperature. Not a datasheet curve at 20 °C — the 20 °C curve tells you almost nothing about the case that matters.
• •The demagnetisation analysis at the combined worst case: maximum temperature and maximum fault current together, not either alone.

If you are evaluating ferrite magnets instead of neodymium — and cost pressure plus rare earth supply risk is pushing some designs that way — the temperature coefficient of coercivity in ferrite is positive. Ferrite gets harder to demagnetise as it warms up.

Which means a ferrite-magnet motor is at its most vulnerable when cold. Every test plan written around neodymium has the temperature backwards for ferrite. A design validated through an Indian summer can fail on a Ladakh winter morning.

The full treatment — including an interactive demagnetisation curve you can drag the temperature on, and the four magnet families compared — is in Magnets and Motors.