Axial Halbach Array Gap Optimization

In a source-free gap, each periodic spatial field component is a solution of Laplace’s equation and decays with distance. For pole pitch tau, the ideal fundamental distance factor is e^(-pi*g/tau); finite geometry, material properties, segmentation, and the return path change the assembled field.

For two matching arrays phased to reinforce at the center plane, the ideal first-harmonic estimate adds their contributions as approximately 2*B1(g/2). This assumes symmetric geometry and does not model mutual material effects, finite edges, stator details, or saturation.

In the ideal periodic model, harmonic n has distance factor e^(-n*pi*g/tau), so higher spatial orders attenuate faster. These transmission factors do not give the source harmonic amplitudes or predict total harmonic distortion, torque ripple, or losses.

There is no universal optimum. Use g/tau to compare ideal field attenuation, then check the required torque or force, mechanical clearance, thermal growth, retention, losses, and cost with models and data for the full machine.

No. The ideal thickness factor 1-e^(-pi*hm/tau) approaches 1 with diminishing returns, while added magnet thickness increases mass and material cost. Evaluate the load line, return path, required field, and supplier data before changing the magnet.

There is no general minimum. Build a worst-case stack from bearing movement, face runout, rotor and stator deflection, thermal growth, magnet and bond-line variation, assembly error, and operating margin for the actual speed and load envelope.

Include non-magnetic layers that sit in the active flux path, such as a face coating or sheet. A radial retention wrap outside the axial path does not add to the axial gap. Use actual dimensions and magnetic properties for each layer.

Axial magnetic attraction, bearing loads, thermal gradients, speed-dependent deformation, and assembly variation can change the gap. Their direction and size depend on the rotor and support design; evaluate the actual structure over its operating envelope rather than applying a generic force or gap correction.

Temperature changes remanence, coercivity, and mechanical dimensions. Coefficients and irreversible-demagnetization limits depend on the exact magnet grade and magnetic circuit. Check supplier curves at the expected temperature and worst-case opposing field.

Measure mechanical clearance with a method suited to the assembly and its access, including runout where relevant. A Hall probe measures magnetic field, not physical clearance; use it only for a separate field check against a defined model or specification.

A simplified model may omit radial edge effects, segment gaps, chamfers, slots, and back-iron saturation. A 2D result does not always overestimate; the difference depends on assumptions and geometry. Compare a representative 3D model with measurements when design decisions require that accuracy.

For a slotted stator, derive any equivalent-gap adjustment from the actual slot opening, tooth pitch, and physical gap. The coefficient is geometry-dependent; detailed slot harmonics and saturation may require a field solution.