Magnetizing Fixture Provides 3-D Flux Flow

George P. Gogue
Joseph J. Stupak, Jr.

G2 Consulting,
Beaverton, OR 97005

A magnetizing fixture produces three-dimensional flux flow for a high-energy magnet ring consisting of eight asymmetric poles of alternating 30° and 60° spans.

Magnetizing fixtures produce the fields required to coerce permanent magnets permanently. These fixtures come in different shapes and sizes determined by the associated permanent magnet. A recent fixture requirement called for magnetization of a high-energy-product magnet ring with eight asymmetric poles and alternating 30° and 60° spans. This fixture required an axial flux path as well as the normal X-Y path. The three dimensional flux flow would enable the fixture to produce a higher magnetizing force than previously possible.

The usual planar approach for magnetizing fixtures would have required a very high magnetizing field. In fact, the field would have been so high that internal heating would make the fixture impossible to use, even if its associated pulse generator could supply the needed power.

This new magnetizer would be used with a bonded neodymium-iron-boron powder magnet requiring a magnetizing intensity of 25,000 oersteds. Therefore, the fixture's required flux-density is higher than the saturation value of any possible steel material for its poles. Nevertheless, the use of steel is very beneficial in fixtures for medium-energy magnet materials. The steel poles greatly reduce the required pulse current, if eddy currents can be kept under control. At even higher required fields, these eddy currents may make it best to do without pole material altogether (Refs 1, 2).

Figure 1. Magnetic Characteristics of Mild Steel.

We decided to use laminated, low-carbon mild steel for the fixture's pole-pieces (laminations reduce eddy currents); Figure 1a shows its magnetization curve. At low flux-densities mild steel has high permeability to magnetic flux (several thousand times higher than air). The permeability decreases in a nonlinear manner at high densities, as shown in Figure 1b. Because the steel saturates above 20,500 gauss, any increase in coercive force produces about the same increase in flux as would be obtained in air (or a vacuum).

In the design of magnetizing fixtures of this type it is sufficient to use a two-part, piece wise linear model. Below 20,000 gauss it is assumed that all the drop in coercive force takes place in the air gaps and magnet, and the steel poles have infinite permeability. The steel's relative permeability compared with air (or a vacuum) is about 40 at this flux-density. The steel saturates above 20,000 gauss, where its assumed permeability is 1. Therefore, we can solve the problem above and below 20,000 gauss, then add the two results for the final answer.

Figure 2. Symmetrical Planar Magnetizing Fixture

In the design of most ring magnet magnetizers for lower-energy product materials (such as ferrites), the magnetizing flux paths are nearly the same as the ones followed in the final application. That is, they are all in one plane (except for end leakage), as shown in Figure 2. Here, the number of poles required on the ring is eight (four pole-pairs). Also shown is an arrangement where two coils, opposite each of its surfaces, magnetize each pole. The two coils produce magnetizing forces in the same direction, so they are additive. However, asymmetry for planar flux paths would require twice as high a field in the 30° poles as the 60° poles. This may not be a problem because the flux levels are still low enough to avoid saturating the steel.

Any increase in field intensity above saturation requires far more coercive force (more ampere-turns) than below saturation. The small amount of increased field strength above saturation needed at the wider poles is not too difficult to achieve (an increase of 5,000 gauss over the 20,000 gauss). However, the increase that would be needed at the short poles in a planar design (30,000 gauss over saturation) is difficult or impossible to achieve in the available volume.

Figure 3. Asymmetrical Magnetizing Fixture

The three-dimensional flux path approach proved to be much better than the one outlined above. As shown in Figures 3a and 3b, a large part of the total flux flows axially. The steel at the center of the fixture is at a significantly different average magnetic potential than that in the steel toward the outside. The flux naturally flows across gaps above and below the magnet-charging region to complete the flux loops. Coercive force "drops" across these gaps are relatively low because of the increased area (compared with the magnet region) and the decreased flux-density.

The condition for magnetization is that everywhere (except at the transitions) the magnetic field must momentarily exceed 25,000 gauss. The maximum gap width is 0.155", which may be used to achieve sufficient accuracy. A more exact calculation would include the small effects of the flux spreading out as it crosses the gap, which is a logarithmic function and a slightly different value.

In air, a vacuum, or many modern magnet materials (including this one), permeability is about 1 gauss/oersted. Up to the point of saturation at 20,000 gauss (20,000 oersteds), the required magnetizing force is:

M1 = (20,000 oersteds)(0.155 in)(2.02 amp-turn/oersted-in) = 6262 ampere-turns................(1)

At higher flux-densities, the effective flux path length is much longer. Using Roter's method (Ref. 3), the path length is about 1.0". An additional 5,000 gauss would require the following magnetizing force:

M2 = (5,000 oersteds) (1.0 in)(2.02 amp-turn/oersted-in = 10,100 ampere-turns.................(2)

Therefore, the total magnetizing force is the sum of equations (1) and (2), or 16,362 ampere-turns. This calculation indicates that 1640A are required to magnetize the magnet material properly with 10 turns of wire around each pole (five inside and five outside the gap). For five turns, the current would have to be 3280A, etc.

When an electric current pulse passes through the magnetizer coils, the resulting magnetic field changes with time. This changing magnetic field across the steel causes eddy currents, which set up magnetic fields that oppose the original field and weaken it. These eddy currents can be large enough to hinder magnetizer operation. Laminated steel poles force eddy currents to remain within individual laminations where they were formed.

Figure 4. Eddy Current Paths in Steel

If the magnetic flux flows in the plane of the lamination rather than normal to it, and if the laminations are thin enough, then current flowing parallel to one lamination face tends to interfere with current flowing along the other face (Figure 4). A lamination thickness calculation requires a determination of the equivalent frequency:


C = Pulse generator capacitance in F

L = Magnetizing fixture inductance in H



P = Permeance

n = Number of turns in coils

Because of steel saturation, permeance (and thus the inductance of the fixture) at peak current is significantly less than that at low current. Either Roter's method or other calculation means can produce the permeance.

At this equivalent frequency (200-2000Hz) you can find "skin depth" from the solution for plane electromagnetic waves impinging on a semi-finite conductive plate (Ref. 4):Equation (5)


l = Radiation wavelength

c = Speed of light

= Electrical conductivity of material

= Magnetic material permeability

In more convenient terms, the skin depth in inches is:


: Relative permeability compared to air (copper = 1)

: Relative resistivity compared to copper (for steel = 6 to 10)

If the lamination thickness is less than two skin depths, you can ignore eddy currents for most purposes. However, the analysis is more complicated for regions where flux paths are normal to the lamination planes. The pulse generator must produce a pulse of sufficient length to allow the required current to pass after the eddy currents have died down, or you must use more resistive high-silicon iron. Another possibility is to cut slots to interrupt the eddy currents.

Besides producing the required level of coercive force for magnetization, you must satisfy other design conditions. For example, the wire temperature must not exceed the breakdown temperature of its insulation, or of the mounting adhesives. Also, if you use the fixture for volume production, it must cool fast enough to allow the required cycle time. In addition, depending on the required use, transitions between magnetic poles in the finished part may have to be very sharp. Furthermore, lead wires and pole pieces may experience large momentary forces; in rare cases forces on the magnets themselves during magnetization must be balanced in some way to prevent destruction of the parts. Moreover, the fixture must be designed for use with a particular pulse generator, based on a knowledge of its characteristics as set in a particular state (if adjustable).

The Need for Magnetizing Fixtures

The manufacturer usually ships unmagnetized magnets and then the OEM magnetizes them either before or after motor assembly. One reason for this is that sometimes magnets cannot hold their fields without surrounding pole pieces of permeable material (normally steel); this depends on both the magnet's size and material. Another reason is that active magnets can attract magnetically permeable dirt that is very difficult to remove. Also, a large volume of active magnets can be dangerous to persons nearby; for example, they could crush fingers or pinch the skin. Furthermore, active magnets can destroy magnetically stored information on tape, disks, or cards, and cause instruments to give false readings, or destroy delicate equipment.

You can sometimes activate weak magnets by other means, but you usually activate them by applying powerful electric pulses to a coil or assembly of coils. The pulses produce a momentary magnetic field strong enough to align the tiny magnetic domains in the magnet (initially the domains point in every direction). When the pulse is over and the applied field disappears, the domains remain aligned and the part is permanently magnetized.

The two components of this process are the pulse generator and the magnetizing fixture. Because of the required magnetizing energy; the pulse generator is usually quite large. The magnetizing fixture may be a coil of wire, or it may be a structure of pole material and various coils.

The magnetic field that coerces the domains into initial alignment is much higher than the resulting field supplied by the magnet afterward, often five to 10 times higher. The magnet manufacturer usually recommends the magnetizing force needed to magnetize the material. This value must be developed by the magnetizing fixture to ensure proper magnetization. If magnets are not driven completely into saturation, their final flux-density levels are both inconsistent and unpredictable.


1. "Magnetizing Techniques for Rare-Earth Magnets," J.R. Place, PCIM July 1988, pp 26-29.

2. "The Analysis of a Magnetizing Fixture for a Multipole Nd-Fe-B Magnet," J.K. Lee, IEEE Trans. on Magnetics, Vol. 24, No. 5, Sept. 1988, pp 2166-71.

3. "Electromagnetic Devices," R.C. Roters, John Wiley & Sons, 1941.

4. Reference Data for Radio Engineers, Howard Sams, ITT, 6th Edition, 1981.