Bonded NdFeB

Specification

BHmax range

7-11 MGOe compression moulded (56-88 kJ/m³); 4-7 MGOe injection moulded (32-56 kJ/m³). Anisotropic bonded grades (HDDR or hot-deformed powder aligned in a field during moulding) reach roughly 15-23 MGOe, at higher powder cost and with an aligning tool.

Remanence (Br)

0.55-0.72 T compression; 0.40-0.60 T injection. Isotropic — magnetisable in any direction, but with no easy axis to exploit. For isotropic material BHmax ≈ Br²/4μ₀, so Br sets the energy product directly.

Intrinsic coercivity (Hcj)

640-1000 kA/m (8-12.5 kOe). Melt-spun powder is inherently high-Hcj because of its nanocrystalline grain structure — demagnetisation resistance is rarely the limiting factor.

Max operating temperature

Binder-limited, not magnet-limited: 120-150 °C epoxy (compression); 120-150 °C polyamide (PA11/PA12 injection); 150-200 °C PPS (injection). Nylon-bound injection grades do not run hotter than epoxy compression grades — PPS is the high-temperature option. Above the binder limit the polymer softens, creeps or degrades before the NdFeB does.

Temperature coefficients

α(Br) ≈ -0.10 to -0.13 %/°C; β(Hcj) ≈ -0.40 %/°C. Reversible loss is comparable to sintered; irreversible loss is usually a binder, creep or powder-oxidation story rather than a coercivity story.

As-moulded tolerance

±0.05 mm on injection-moulded features; ±0.05-0.1 mm typical compression, ±0.1 mm on larger dimensions — achievable without grinding. Ground sintered faces hold ±0.02-0.05 mm, so tolerance alone is not the argument: it is holding a usable tolerance on a net-shape multipole geometry in one operation.

Size and wall section

Rings, arcs, discs and cups from roughly 3 mm to 80 mm OD. Wall down to ~1 mm compression, ~0.6 mm injection. Insert-moulding onto shafts and hubs is routine.

Coating and density

Epoxy spray/e-coat, impregnation or parylene for compression grades; injection-moulded parts often run bare in dry service. Density 4.2-5.0 g/cm³ injection and ~5.8-6.2 g/cm³ compression, against 7.4-7.6 for sintered. The polymer isolates each particle and removes the continuous Nd-rich grain-boundary network that lets sintered NdFeB corrode intergranularly — but bonded parts retain residual porosity, take up moisture, and surface-exposed particles still oxidise. Coat for humid or salt-laden service.

Typical applications

• BLDC and stepper rotors for pumps, fans and HVAC actuators — thin multipole rings insert-moulded straight onto the shaft, removing a separate bond-and-balance operation
• Position and speed sensor rings for encoders, ABS wheel sensors and steering-angle sensors, where pole-pitch uniformity matters far more than peak flux density
• Small precision motors in office automation, HDD spindle and camera focus mechanisms, where thin walls and tight as-moulded tolerances rule out ground sintered parts
• Multipole couplings and small magnetic gears in dosing and metering equipment, where the moulded polymer skin tolerates chemical exposure that would attack the coating on a plated sintered magnet — note that torque-dense pump couplings usually still go sintered
• Complex-geometry parts — cups, arcs with integral features, over-moulded assemblies — where sintering plus grinding costs more than the energy product is worth
• Consumer and appliance actuators, latches and small linear motors that need multipole patterning at low piece price rather than maximum torque density

Engineering notes

• The energy product penalty is real and unavoidable. At 8 MGOe against an N42 sintered part you need roughly 5x the magnet volume for the same stored airgap energy; for the same airgap flux the multiplier is set by the load line and the Br ratio (about 2x on Br alone), so run the magnetic circuit rather than scaling a single number. If the envelope is tight and the geometry is simple, sintered wins on cost per unit of flux — bonded only pays when the shape or the pole count is doing work for you.
• The binder sets the thermal ceiling, and it fails differently from a magnet. Sintered NdFeB above its limit loses flux irreversibly but stays a solid part; a bonded magnet above its binder limit can creep, swell or crack, and an insert-moulded rotor can lose its grip on the shaft. Qualify at the real hot-spot temperature, not the nominal ambient — and check which binder you are actually being quoted, since PA12 and PPS are 50 °C apart.
• Mechanical strength is modest. Flexural strength runs roughly 40-80 MPa depending on binder and loading, against ~250 MPa transverse rupture strength for sintered. Bonded parts tolerate handling and light press-fits better than sintered (they do not chip), but they will not carry structural load, and high-speed rotors usually still need a retaining sleeve or a designed-in hub.
• Tooling is a real up-front cost and a real lead time. Compression tools are cheaper; injection tools for multipole rings are not, and neither is the magnetising fixture. Anisotropic bonded grades need an alignment field in the tool, which raises the tooling cost again. Below a few thousand pieces a year the amortised tooling often erases the savings from skipping grinding — run the numbers before committing.
• Long-term moisture uptake is the quiet failure mode. The polymer removes the intergranular corrosion path but is not a moisture barrier, and the part is not fully dense; nylon binders in particular absorb water, which swells the part and slowly oxidises exposed particles. For humid, salt-spray or automotive underbonnet service, specify a coating and ask for 85 °C/85% RH data rather than assuming the binder handles it.