How Do Heavy Rare Earth Free NdFeB Magnets Actually Perform in EV Traction Motors?

Why Procurement and Design Engineers Must Reconsider Dy/Tb-Free Magnet Grades

The decision facing an EV traction-motor engineer is no longer only “which magnet grade,” but which grade remains viable once rare-earth supply risk and customer compliance requirements are priced into the design. For more than a decade, dysprosium and terbium have been the usual additives that help sintered NdFeB retain coercivity at the elevated temperatures found in a traction-motor rotor. Those two elements also sit in a concentrated, geopolitically sensitive part of the supply chain, and they are the elements that a growing set of customer specifications now seeks to reduce or eliminate. The engineering question is therefore whether the coercivity and temperature-headroom penalty of a heavy-rare-earth-free (HRE-free) grade is acceptable against the compliance, cost-stability, and supply-security benefits it can buy. That is a trade-off between a measurable drop in high-temperature coercivity margin and a reduction in heavy-rare-earth exposure that cannot be read off a BH curve.

In this article, “HRE-free” means a sintered NdFeB magnet whose coercivity is obtained without dysprosium or terbium — typically by grain-boundary diffusion of light rare earths, by grain-boundary-phase engineering, by higher neodymium/praseodymium content with a refined microstructure, or by a combination of these routes. The performance question is not whether such magnets can function in a vehicle — production use already exists — but whether a given grade holds its working point across the torque–speed–temperature envelope for which your motor is actually rated.

HRE-Free and Dy-Containing Grades Diverge on the Axes That Drive Traction-Motor Design

The comparison below does not use a generic peak-(BH)max table. For a traction motor the deciding factors are coercivity margin at temperature, the temperature at which the grade still delivers usable flux without irreversible loss, the coercivity-processing route, supply-chain exposure, and cost trajectory. Grade letters such as UH and EH are room-temperature coercivity / conventional maximum-use classes; they are not a substitute for a demagnetization curve at the rotor’s peak temperature.

The engineering takeaway is straightforward: an HRE-free grade does not automatically reduce torque, but it usually reduces temperature headroom and knee margin. That is the axis on which rotor thermal design, magnet thickness, and peak-current demagnetization must be re-examined.

Coercivity Margin Shrinks with Temperature and Must Be Budgeted in the Magnetic Circuit

Intrinsic coercivity of sintered NdFeB falls with temperature. Over the range of interest for a liquid-cooled traction rotor, a linear temperature coefficient is a common first-order engineering approximation; it is not a substitute for a measured J–H curve at the peak magnet temperature. What decides survival is not Hcj at 20 °C, and not Hcj(T) compared with zero. It is whether the local operating point remains above the knee of the demagnetization curve at that temperature.

The working point is the intersection of the magnet’s J–H (or B–H) curve at temperature T with the load line of the magnetic circuit. That load line is set by geometry — magnet thickness in the magnetization direction, air gap, leakage, and surrounding iron — often summarized as a permeance coefficient, plus the demagnetizing field from stator armature reaction at peak current. A useful design inequality is therefore:

with a margin that covers hot-spot scatter, build tolerance, and curve uncertainty. Here $H_d$ is the local demagnetizing field in the magnet and $H_k(T)$ is the knee field at temperature T, not the intrinsic coercivity $H_{\mathrm{c}\mathrm{j}}(T)$. $H_{\mathrm{c}\mathrm{j}}$ is the field at which polarization has already fallen to zero; designing to $H_{\mathrm{c}\mathrm{j}}$ is designing after irreversible loss has begun.

Two geometry facts are easy to get backwards. In an assembled motor, a thinner magnet (smaller length in the magnetization direction) generally lowers the permeance coefficient and increases self-demagnetizing field. Thickening the magnet, reducing the gap, or otherwise raising permeance coefficient improves demagnetization withstand; thinning the magnet to “reduce N” does not. Armature reaction at peak torque is often the larger threat in a traction motor, especially with concentrated windings, and it is local — IPM corners and SPM trailing edges are typical first-to-fail regions. A single ellipsoid demagnetizing factor, or a 2-D load line that ignores peak current, will not capture that.

If a Dy/Tb-containing grade is replaced by an HRE-free grade of the same nominal temperature class, torque at a given current need not fall when Br is comparable, but coercivity margin usually does. That margin has to be recovered by one or more of the following levers:

• a lower peak magnet temperature (cooling, loss control, or a less severe duty);
• a higher permeance coefficient (typically more magnet thickness and/or a smaller gap — not a thinner magnet);
• a lower peak demagnetizing MMF from the stator (current limit, control strategy, or circuit geometry that protects the magnet corners);
• a grade whose measured high-temperature knee actually clears the local $H_d$.

Any of these can cost mass, efficiency, current capability, or cooling capacity. The decision is a system trade-off, not a drop-in grade swap.

A second quantitative effect is the temperature at which irreversible flux loss begins for your load line. That temperature is not a linear extrapolation from a 20 °C datasheet, and it is not guaranteed by the UH/EH letter. It depends on knee shape, grain-boundary processing, and the actual demagnetizing field, including armature reaction. For an HRE-free candidate the onset is often earlier than for a Dy/Tb-containing grade of similar 20 °C class. The only reliable evidence is a demagnetization curve measured at the peak magnet temperature you will actually run — and, where the duty includes high current at that temperature, a check of the local operating point, not only the open-circuit hot curve. Insist on that data before the rotor geometry is frozen.

An HRE-Free Grade Is Viable Only Inside a Defined Temperature, Load-Line, and Duty Window

Treat an HRE-free grade as a valid choice when the following window is satisfied, and as the wrong choice when the window is exceeded.

First, peak magnet temperature must be known from a thermal map, not assumed from coolant temperature, and held below the temperature at which the measured knee is reached on the actual load line. Datasheet “maximum operating temperature” already assumes a reference permeance coefficient (often near unity) and does not include your hot spots or peak-current armature reaction. A practical starting rule is to keep peak magnet temperature, including local hot spots, on the order of 20–30 °C below the datasheet Tmax of the grade, then confirm that the hot curve still clears the load line at peak current. The 20–30 °C band is a margin for measurement uncertainty and rotor variation; it is not a physical constant. If cooling cannot hold a verified margin, the HRE-free grade is the wrong choice regardless of Dy/Tb savings.

Second, the demagnetizing field must be controlled in the real circuit. Magnet thickness, air gap, pole geometry, and peak d-axis current determine whether the operating point sits above the knee. If the existing design used a thin magnet and relied on a high-Hcj Dy/Tb grade to survive peak current, switching to HRE-free is a geometry and control problem, not a material substitution.

Third, the worst-case operating point must be defined honestly: peak torque at peak current at peak magnet temperature, including transients the inverter will actually allow. Qualification at rated torque and a mean duty cycle will miss the event that pushes the magnet through the knee.

Fourth, parasitic rotor heating has to be treated as a direct attack on the coercivity budget. Segmented magnets, attention to magnet eddy currents under PWM and space-harmonic fields, and a rotor thermal path that does not dump extra heat into the magnet are part of making an HRE-free grade work. They are not optional accessories.

Fifth, where coercivity depends on grain-boundary diffusion, magnet thickness cuts both ways: thicker magnets help permeance coefficient, but diffusion may not be uniform through a thick part. Diffusion quality, not the certificate’s class letter, sets the high-temperature knee. That is a process and inspection issue, and it belongs in the sourcing specification.

HRE-Free Magnets Should Not Be Used When the Knee, the Thermal Limit, or the Duty Cycle Cannot Be Controlled

The most common failure mode is not a single catastrophic demagnetization event but cumulative irreversible flux loss. When the magnet operates near its knee, thermal and current cycles can each take a small permanent increment of remanence. Over a large number of cycles, torque drifts below specification. This mode is easy to miss in a short qualification test because it appears as a slow drift rather than a hard fault.

A second failure mode is a misapplied temperature rating. If the grade is sold on room-temperature Hcj while grain-boundary processing is marginal, the real knee at 150–180 °C can sit well below what the class letter implies. A datasheet class is not a motor specification. Require the demagnetization curve at the actual peak magnet temperature and verify the knee against the local load line, including peak current.

A third boundary is the application envelope itself. HRE-free magnets are generally the wrong choice when the motor routinely runs high sustained overload, when rotor cooling cannot be improved enough to protect the knee, or when peak magnet temperature is genuinely above what the hot curve supports and the geometry cannot be changed. In those cases a Dy/Tb-containing grade remains the physically correct choice; cost and supply arguments are secondary to keeping the magnet above the knee.

A fourth boundary is a compliance-driven substitution with no redesign. If a customer mandates “no heavy rare earths,” the commercial trade-off disappears and the motor must be designed around the HRE-free grade — with temperature, thickness, peak current, and inspection of hot curves fully engineered. Substituting the grade into an unchanged Dy/Tb rotor and hoping the margins still hold is how the slow flux-loss mode appears in the field.

AIC Engineering Fits HRE-Free Magnets to the Motor Envelope Rather Than Swapping a Grade in Isolation

The value of an application partner in this decision is not the grade letter on a carton. It is verification that the grade survives the specific operating envelope. AIC Engineering’s magnetic-circuit and magnet-application structure design work is aimed at that check: the load line, local demagnetizing field, and thermal map have to be taken at the design stage, not after the rotor is tooled. The AIC Engineering team can model the magnet working point against pole geometry, current profile, and temperature distribution, and recommend whether an HRE-free grade is viable or whether thickness, cooling, or circuit geometry must change to recover knee margin.

For the transition phase, AIC Engineering’s rapid prototyping, with samples available in as little as 3–7 days, allows an HRE-free candidate to be checked on a rotor or a test fixture before production tooling is committed. That is the practical way to answer whether the grade holds its working point in this motor.

Because the failure modes above are measurable, permanent-magnet quality inspection is the third element that matters. Demagnetization curves at elevated temperature, coercivity verification against the specified class, and dimensional and coating checks are what separate a grade that works on paper from one that works in the field. Inspection is aligned with the industry measurement methods used for permanent magnets, so the data can be compared directly with what the motor design team needs.

For a procurement team that is weighing supply security next to performance, AIC Engineering’s global supply and regional delivery support is part of the same risk reduction. An HRE-free grade is often chosen to reduce exposure to a concentrated heavy-rare-earth chain; consistent quality across lots and delivery from more than one region is what makes that choice operational rather than rhetorical.

Next Steps Convert the Grade Choice into a Verified Working Point

1. Extract the real operating envelope. Define peak magnet temperature (including hot spots), peak torque at peak current, and the worst-case transient the inverter will allow, and confirm these against the cooling design before comparing grades.
2. Request a demagnetization curve at temperature, not only a class letter. Ask for the J–H or B–H curve at the actual peak magnet temperature and verify the knee against the calculated local load line, including armature reaction.
3. Quantify knee margin before switching. Compare the present Dy/Tb-containing grade and the HRE-free candidate on the same geometry and the same peak-current, peak-temperature point. Treat a 20–30 °C buffer below datasheet Tmax as a starting thermal margin, then confirm it on the hot curve.
4. Contact AIC Engineering for a consultation and a magnetic-circuit review. The team can model the magnet working point, recommend an HRE-free grade and geometry that fit the envelope, and deliver a rapid prototype in 3–7 days for bench validation. Visit https://www.aicmagnetics.com to start the discussion on the traction-motor magnet design.

References

• Gutfleisch, O., Willard, M. A., Brück, E., Chen, C. H., Sankar, S. G., & Liu, J. P. (2011). Magnetic materials and devices for the 21st century: Stronger, lighter, and more energy efficient. Advanced Materials, 23(7), 821–842.
• Coey, J. M. D. (2002). Permanent magnet applications. Journal of Magnetism and Magnetic Materials, 248(3), 441–456.
• Hirota, K., Nakamura, H., Minowa, T., & Honshima, M. (2006). Coercivity enhancement by the grain boundary diffusion process to Nd–Fe–B sintered magnets. IEEE Transactions on Magnetics, 42(10), 2909–2911.
• European Commission. (2023). Study on the Critical Raw Materials for the EU 2023. Directorate-General for Internal Market, Industry, Entrepreneurship and SMEs.
• Standardization Administration of China. (2017). GB/T 13560-2017 Sintered neodymium iron boron permanent magnets.
• IEC 60404-5:2015. Magnetic materials — Part 5: Permanent magnet (magnetically hard) materials — Methods of measurement of magnetic properties. International Electrotechnical Commission.